{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "59e1074b-b7a6-4704-b2a9-c42fddcd75c1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "10b38124-00be-4a49-8d84-91b92f8d950f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Destination Port</th>\n",
       "      <th>Flow Duration</th>\n",
       "      <th>Total Fwd Packets</th>\n",
       "      <th>Total Backward Packets</th>\n",
       "      <th>Total Length of Fwd Packets</th>\n",
       "      <th>Total Length of Bwd Packets</th>\n",
       "      <th>Fwd Packet Length Max</th>\n",
       "      <th>Fwd Packet Length Min</th>\n",
       "      <th>Fwd Packet Length Mean</th>\n",
       "      <th>Fwd Packet Length Std</th>\n",
       "      <th>...</th>\n",
       "      <th>min_seg_size_forward</th>\n",
       "      <th>Active Mean</th>\n",
       "      <th>Active Std</th>\n",
       "      <th>Active Max</th>\n",
       "      <th>Active Min</th>\n",
       "      <th>Idle Mean</th>\n",
       "      <th>Idle Std</th>\n",
       "      <th>Idle Max</th>\n",
       "      <th>Idle Min</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>80</td>\n",
       "      <td>68855579</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>1038</td>\n",
       "      <td>11595</td>\n",
       "      <td>346</td>\n",
       "      <td>0</td>\n",
       "      <td>103.800000</td>\n",
       "      <td>167.133879</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>998.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>998</td>\n",
       "      <td>998</td>\n",
       "      <td>6.830000e+07</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>68300000</td>\n",
       "      <td>68300000</td>\n",
       "      <td>DoS Hulk</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>53</td>\n",
       "      <td>196</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>70</td>\n",
       "      <td>174</td>\n",
       "      <td>35</td>\n",
       "      <td>35</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>123</td>\n",
       "      <td>118</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>96</td>\n",
       "      <td>96</td>\n",
       "      <td>48</td>\n",
       "      <td>48</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>20</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>80</td>\n",
       "      <td>295657</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>1114</td>\n",
       "      <td>15841</td>\n",
       "      <td>1084</td>\n",
       "      <td>0</td>\n",
       "      <td>159.142857</td>\n",
       "      <td>407.829796</td>\n",
       "      <td>...</td>\n",
       "      <td>20</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>80</td>\n",
       "      <td>15705351</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>407</td>\n",
       "      <td>452</td>\n",
       "      <td>377</td>\n",
       "      <td>0</td>\n",
       "      <td>58.142857</td>\n",
       "      <td>140.620563</td>\n",
       "      <td>...</td>\n",
       "      <td>20</td>\n",
       "      <td>360718.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>360718</td>\n",
       "      <td>360718</td>\n",
       "      <td>9.767208e+06</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>9767208</td>\n",
       "      <td>9767208</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783351</th>\n",
       "      <td>41272</td>\n",
       "      <td>98689782</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>11595</td>\n",
       "      <td>327</td>\n",
       "      <td>5792</td>\n",
       "      <td>0</td>\n",
       "      <td>1449.375000</td>\n",
       "      <td>2046.673464</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>13859.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>13859</td>\n",
       "      <td>13859</td>\n",
       "      <td>9.860000e+07</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>98600000</td>\n",
       "      <td>98600000</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783352</th>\n",
       "      <td>443</td>\n",
       "      <td>90956783</td>\n",
       "      <td>63</td>\n",
       "      <td>77</td>\n",
       "      <td>1041</td>\n",
       "      <td>135313</td>\n",
       "      <td>373</td>\n",
       "      <td>0</td>\n",
       "      <td>16.523810</td>\n",
       "      <td>68.018939</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>198255.7778</td>\n",
       "      <td>362537.3486</td>\n",
       "      <td>1165022</td>\n",
       "      <td>77291</td>\n",
       "      <td>9.908053e+06</td>\n",
       "      <td>290822.6482</td>\n",
       "      <td>10000000</td>\n",
       "      <td>9132848</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783353</th>\n",
       "      <td>443</td>\n",
       "      <td>1241813</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>28</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783354</th>\n",
       "      <td>53</td>\n",
       "      <td>71509</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>78</td>\n",
       "      <td>330</td>\n",
       "      <td>39</td>\n",
       "      <td>39</td>\n",
       "      <td>39.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>32</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783355</th>\n",
       "      <td>4306</td>\n",
       "      <td>2167</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>20</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>BENIGN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1782497 rows × 79 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          Destination Port   Flow Duration   Total Fwd Packets  \\\n",
       "0                       80        68855579                  10   \n",
       "1                       53             196                   2   \n",
       "2                      123             118                   2   \n",
       "3                       80          295657                   7   \n",
       "4                       80        15705351                   7   \n",
       "...                    ...             ...                 ...   \n",
       "1783351              41272        98689782                   8   \n",
       "1783352                443        90956783                  63   \n",
       "1783353                443         1241813                   3   \n",
       "1783354                 53           71509                   2   \n",
       "1783355               4306            2167                   2   \n",
       "\n",
       "          Total Backward Packets  Total Length of Fwd Packets  \\\n",
       "0                              6                         1038   \n",
       "1                              2                           70   \n",
       "2                              2                           96   \n",
       "3                             10                         1114   \n",
       "4                              5                          407   \n",
       "...                          ...                          ...   \n",
       "1783351                        5                        11595   \n",
       "1783352                       77                         1041   \n",
       "1783353                        3                            0   \n",
       "1783354                        2                           78   \n",
       "1783355                        0                            0   \n",
       "\n",
       "          Total Length of Bwd Packets   Fwd Packet Length Max  \\\n",
       "0                               11595                     346   \n",
       "1                                 174                      35   \n",
       "2                                  96                      48   \n",
       "3                               15841                    1084   \n",
       "4                                 452                     377   \n",
       "...                               ...                     ...   \n",
       "1783351                           327                    5792   \n",
       "1783352                        135313                     373   \n",
       "1783353                             0                       0   \n",
       "1783354                           330                      39   \n",
       "1783355                             0                       0   \n",
       "\n",
       "          Fwd Packet Length Min   Fwd Packet Length Mean  \\\n",
       "0                             0               103.800000   \n",
       "1                            35                35.000000   \n",
       "2                            48                48.000000   \n",
       "3                             0               159.142857   \n",
       "4                             0                58.142857   \n",
       "...                         ...                      ...   \n",
       "1783351                       0              1449.375000   \n",
       "1783352                       0                16.523810   \n",
       "1783353                       0                 0.000000   \n",
       "1783354                      39                39.000000   \n",
       "1783355                       0                 0.000000   \n",
       "\n",
       "          Fwd Packet Length Std  ...   min_seg_size_forward  Active Mean  \\\n",
       "0                    167.133879  ...                     32     998.0000   \n",
       "1                      0.000000  ...                     32       0.0000   \n",
       "2                      0.000000  ...                     20       0.0000   \n",
       "3                    407.829796  ...                     20       0.0000   \n",
       "4                    140.620563  ...                     20  360718.0000   \n",
       "...                         ...  ...                    ...          ...   \n",
       "1783351             2046.673464  ...                     32   13859.0000   \n",
       "1783352               68.018939  ...                     32  198255.7778   \n",
       "1783353                0.000000  ...                     28       0.0000   \n",
       "1783354                0.000000  ...                     32       0.0000   \n",
       "1783355                0.000000  ...                     20       0.0000   \n",
       "\n",
       "          Active Std   Active Max   Active Min     Idle Mean     Idle Std  \\\n",
       "0             0.0000          998          998  6.830000e+07       0.0000   \n",
       "1             0.0000            0            0  0.000000e+00       0.0000   \n",
       "2             0.0000            0            0  0.000000e+00       0.0000   \n",
       "3             0.0000            0            0  0.000000e+00       0.0000   \n",
       "4             0.0000       360718       360718  9.767208e+06       0.0000   \n",
       "...              ...          ...          ...           ...          ...   \n",
       "1783351       0.0000        13859        13859  9.860000e+07       0.0000   \n",
       "1783352  362537.3486      1165022        77291  9.908053e+06  290822.6482   \n",
       "1783353       0.0000            0            0  0.000000e+00       0.0000   \n",
       "1783354       0.0000            0            0  0.000000e+00       0.0000   \n",
       "1783355       0.0000            0            0  0.000000e+00       0.0000   \n",
       "\n",
       "          Idle Max   Idle Min     Label  \n",
       "0         68300000   68300000  DoS Hulk  \n",
       "1                0          0    BENIGN  \n",
       "2                0          0    BENIGN  \n",
       "3                0          0    BENIGN  \n",
       "4          9767208    9767208    BENIGN  \n",
       "...            ...        ...       ...  \n",
       "1783351   98600000   98600000    BENIGN  \n",
       "1783352   10000000    9132848    BENIGN  \n",
       "1783353          0          0    BENIGN  \n",
       "1783354          0          0    BENIGN  \n",
       "1783355          0          0    BENIGN  \n",
       "\n",
       "[1782497 rows x 79 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('Train_ULAK.csv').dropna()\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "5ddc80df-7dcd-45a9-920b-4a34f0fd9b08",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index([' Destination Port', ' Flow Duration', ' Total Fwd Packets',\n",
       "       ' Total Backward Packets', 'Total Length of Fwd Packets',\n",
       "       ' Total Length of Bwd Packets', ' Fwd Packet Length Max',\n",
       "       ' Fwd Packet Length Min', ' Fwd Packet Length Mean',\n",
       "       ' Fwd Packet Length Std', 'Bwd Packet Length Max',\n",
       "       ' Bwd Packet Length Min', ' Bwd Packet Length Mean',\n",
       "       ' Bwd Packet Length Std', 'Flow Bytes/s', ' Flow Packets/s',\n",
       "       ' Flow IAT Mean', ' Flow IAT Std', ' Flow IAT Max', ' Flow IAT Min',\n",
       "       'Fwd IAT Total', ' Fwd IAT Mean', ' Fwd IAT Std', ' Fwd IAT Max',\n",
       "       ' Fwd IAT Min', 'Bwd IAT Total', ' Bwd IAT Mean', ' Bwd IAT Std',\n",
       "       ' Bwd IAT Max', ' Bwd IAT Min', 'Fwd PSH Flags', ' Bwd PSH Flags',\n",
       "       ' Fwd URG Flags', ' Bwd URG Flags', ' Fwd Header Length',\n",
       "       ' Bwd Header Length', 'Fwd Packets/s', ' Bwd Packets/s',\n",
       "       ' Min Packet Length', ' Max Packet Length', ' Packet Length Mean',\n",
       "       ' Packet Length Std', ' Packet Length Variance', 'FIN Flag Count',\n",
       "       ' SYN Flag Count', ' RST Flag Count', ' PSH Flag Count',\n",
       "       ' ACK Flag Count', ' URG Flag Count', ' CWE Flag Count',\n",
       "       ' ECE Flag Count', ' Down/Up Ratio', ' Average Packet Size',\n",
       "       ' Avg Fwd Segment Size', ' Avg Bwd Segment Size',\n",
       "       ' Fwd Header Length.1', 'Fwd Avg Bytes/Bulk', ' Fwd Avg Packets/Bulk',\n",
       "       ' Fwd Avg Bulk Rate', ' Bwd Avg Bytes/Bulk', ' Bwd Avg Packets/Bulk',\n",
       "       'Bwd Avg Bulk Rate', 'Subflow Fwd Packets', ' Subflow Fwd Bytes',\n",
       "       ' Subflow Bwd Packets', ' Subflow Bwd Bytes', 'Init_Win_bytes_forward',\n",
       "       ' Init_Win_bytes_backward', ' act_data_pkt_fwd',\n",
       "       ' min_seg_size_forward', 'Active Mean', ' Active Std', ' Active Max',\n",
       "       ' Active Min', 'Idle Mean', ' Idle Std', ' Idle Max', ' Idle Min',\n",
       "       ' Label'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "85de1086-3e8e-498a-b965-8481cbef0f4f",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Let's see if some assumption based on domain knowlwdge hold\n",
    "# For this reason, let's split by label to start\n",
    "gb=df.groupby(' Label')\n",
    "dfs = [gb.get_group(x) for x in gb.groups]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6e8d0896-6b29-469a-8281-c24825a3dddc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['BENIGN'], dtype=object),\n",
       " array(['Bot'], dtype=object),\n",
       " array(['DDoS'], dtype=object),\n",
       " array(['DoS GoldenEye'], dtype=object),\n",
       " array(['DoS Hulk'], dtype=object),\n",
       " array(['DoS Slowhttptest'], dtype=object),\n",
       " array(['DoS slowloris'], dtype=object),\n",
       " array(['FTP-Patator'], dtype=object),\n",
       " array(['Heartbleed'], dtype=object),\n",
       " array(['Infiltration'], dtype=object),\n",
       " array(['PortScan'], dtype=object),\n",
       " array(['SSH-Patator'], dtype=object),\n",
       " array(['Web Attack � Brute Force'], dtype=object),\n",
       " array(['Web Attack � Sql Injection'], dtype=object),\n",
       " array(['Web Attack � XSS'], dtype=object)]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "# This will let us know how to associate each group with a label\n",
    "labels_per_group = [np.unique(df[[' Label']]) for df in dfs]\n",
    "labels_per_group"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cca72042-0680-4901-8bc2-8546d0c28995",
   "metadata": {},
   "source": [
    "# Level 1 filtering, the type of attack can only happen on particular destination port"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ccb7bb25-f193-4f29-a904-9c33691ac241",
   "metadata": {},
   "source": [
    "## Let's check if web attacks happen only on ports 80 and 443"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "22b2d4a7-57e3-4eb5-826b-42c50e0f3b5d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['Web Attack � Brute Force'], dtype=object),\n",
       " array(['Web Attack � Sql Injection'], dtype=object),\n",
       " array(['Web Attack � XSS'], dtype=object)]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Confirm the indeces for web attacks\n",
    "labels_per_group[12:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3f743d5a-9607-466f-8fb7-8f067c94ee53",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([80]), array([80]), array([80])]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# In the training set only port 80 shows up\n",
    "[np.unique(df[[' Destination Port']]) for df in dfs[12:]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96f7c14e-466d-4e7e-ad96-073bf67d23ee",
   "metadata": {},
   "source": [
    "## Lets check that SSH_Patator only happens on port 22"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "68c06679-c9ae-4dfd-b682-8302611d33ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['SSH-Patator'], dtype=object)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Confirm the indeces for SSH-Patator\n",
    "labels_per_group[11]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e7e742cd-b8dc-4c88-8524-8df7e674dfac",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([22])]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# In the training set only port 22 shows up\n",
    "[np.unique(df[[' Destination Port']]) for df in dfs[11:12]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f939d6e-50e9-48ca-9358-90370d53661a",
   "metadata": {},
   "source": [
    "## Let's check the FTP-Patator only happens on port 21"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b2be1552-a3f3-43b7-82ab-98b984cc03f8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['FTP-Patator'], dtype=object)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Confirm the indeces for FTP-Patator\n",
    "labels_per_group[7]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "25b953dd-2ab0-413a-aeb2-a50d024b6873",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([21])]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# In the training set only port 21 shows up\n",
    "[np.unique(df[[' Destination Port']]) for df in dfs[7:8]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a6475a5-8e4f-4b77-8374-d25c7a3050e9",
   "metadata": {},
   "source": [
    "# Let's start on Level 2 filter"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb3b92c2-1e43-43c6-b7c5-8096fc94a67d",
   "metadata": {},
   "source": [
    "## See how well the following rule works: \n",
    "\n",
    "BENIGN: Normal traffic\n",
    "\n",
    "if ['Destination Port'] in [80, 443] and ['Flow Duration'] < threshold:\n",
    "    return 'BENIGN'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b50aac85-caa4-48dd-bd57-c1625ea04e15",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['BENIGN'], dtype=object)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Confirm the indeces for Benign\n",
    "labels_per_group[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "292ce76a-55a3-4a35-9688-20b3ee5c21a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "bdf = dfs[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c4f9b82c-6685-4b51-9270-e75bcd3e7028",
   "metadata": {},
   "outputs": [],
   "source": [
    "p80fd = bdf[bdf[' Destination Port'] == 80][[' Flow Duration']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "60d73cf6-be46-4ec9-87a8-8340477d1709",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " Flow Duration    2.764066e+07\n",
       "dtype: float64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p80fd.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "8cda607c-e0cc-4778-8bcd-acd41d9b0fb2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " Flow Duration    119999979\n",
       "dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p80fd.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "66b6c067-82ea-4f72-bd56-f8e56af2095a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " Flow Duration    4.324591e+07\n",
       "dtype: float64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p80fd.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "28ff3956-ab89-43ad-af31-fd807a154e87",
   "metadata": {},
   "outputs": [],
   "source": [
    "np80=dfs[1:]\n",
    "np80fd = [df[df[' Destination Port'] == 80][' Flow Duration'] for df in np80]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "ac02c865-2c34-45dc-a02f-b14f0c4774b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[nan,\n",
       " 17073736.94560649,\n",
       " 23386412.162399754,\n",
       " 57345934.07025991,\n",
       " 57551770.44659353,\n",
       " 57075233.23938647,\n",
       " nan,\n",
       " nan,\n",
       " nan,\n",
       " 739167.1764705882,\n",
       " nan,\n",
       " 6638958.354056902,\n",
       " 2513092.8333333335,\n",
       " 7001995.16097561]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[df.mean() for df in np80fd]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "25e46846-e151-446f-94a6-a11aeddb9518",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[nan,\n",
       " 103941558,\n",
       " 119311937,\n",
       " 119611868,\n",
       " 119800736,\n",
       " 119046064,\n",
       " nan,\n",
       " nan,\n",
       " nan,\n",
       " 9605093,\n",
       " nan,\n",
       " 35451960,\n",
       " 5086516,\n",
       " 70203058]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[df.max() for df in np80fd]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "aad28091-45eb-4ade-892d-ab3088ad8270",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[nan,\n",
       " 31142536.391154718,\n",
       " 27532239.30393215,\n",
       " 45965072.766122654,\n",
       " 36274513.68910278,\n",
       " 49442828.4695022,\n",
       " nan,\n",
       " nan,\n",
       " nan,\n",
       " 1930486.2004055232,\n",
       " nan,\n",
       " 6637697.083727899,\n",
       " 2624856.174791534,\n",
       " 10491313.551777482]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[df.std() for df in np80fd]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "d01cdcc4-3fff-4c11-8d40-82f2995ad155",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Create the plot\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "# Plotting for 'BENIGN'\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(p80fd[' Flow Duration'], bins=50, alpha=0.5, label='BENIGN')  # Note that p80fd is a DataFrame with one column\n",
    "plt.title(' Flow Duration under Destination Port 80 for BENIGN')\n",
    "plt.xlabel('Flow Duration')\n",
    "plt.ylabel('Frequency')\n",
    "\n",
    "# Plotting for non-'BENIGN'\n",
    "plt.subplot(1, 2, 2)\n",
    "for i, df in enumerate(np80fd):  # np80fd is a list of Series\n",
    "    plt.hist(df, bins=50, alpha=0.5, label=labels_per_group[i+1])  # assuming labels_per_group[0] is 'BENIGN'\n",
    "plt.title(' Flow Duration under Destination Port 80 for NON-BENIGN')\n",
    "plt.xlabel('Flow Duration')\n",
    "plt.ylabel('Frequency')\n",
    "\n",
    "plt.legend(loc='upper right')\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25360415-0f04-44d2-b6b5-68d9d1568561",
   "metadata": {},
   "source": [
    "Incomplete Data: Many non-BENIGN classes have nan as their statistics, which might indicate a lack of samples. In such cases, the heuristic might be even more prone to errors.\n",
    "\n",
    "Port Limitation: Using only ports 80 and 443 might ignore many other kinds of benign traffic that don't use these ports.\n",
    "\n",
    "Suggestions for Machine Learning Model:\n",
    "Random Forest\n",
    "Argument: Random Forests handle high dimensionality and feature interaction well, making them an excellent choice for complex datasets with mixed types of variables. They also provide feature importance metrics, which can be valuable for interpretability.\n",
    "Based on Statistics: The statistics show a high level of variability among features. Random Forest can capture these without assuming any underlying data distribution.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2d7b68d-02d8-4c52-be31-571cb380ecfd",
   "metadata": {},
   "source": [
    "## See how well the following rule works: \n",
    "\n",
    "Dos Hulk\n",
    "\n",
    "if ['Fwd Packet Length Max'] > threshold and ['Flow Packets/s'] > threshold:\n",
    "    return 'DoS Hulk'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "0e027b28-f687-4f20-a3b4-e6d54b304c56",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['DDoS'], dtype=object)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Confirm the indeces for DoS Hulk\n",
    "labels_per_group[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "612ae717-517e-41ec-92ca-31cdb692d8b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "ddosdf= dfs[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "fee14c4b-766b-4737-b44f-5afff86aedf9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " Fwd Packet Length Max    14.932256\n",
       "dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ddosdf[[' Fwd Packet Length Max']].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "e1efd21b-3ef7-4f1e-8841-c94170e8107c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " Fwd Packet Length Max    20\n",
       "dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ddosdf[[' Fwd Packet Length Max']].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "c25d32c8-6f7b-42ba-848b-247a3fc1e9e2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " Fwd Packet Length Max    6.728072\n",
       "dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ddosdf[[' Fwd Packet Length Max']].std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "73714be1-6af5-48d2-a795-94b0fd5c9dc3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'DoS Hulk' is at index 4 in labels_per_group\n",
      "Statistics for 'Fwd Packet Length Max' under 'DoS Hulk'\n",
      "Mean: 233.66139223043814\n",
      "Max: 423\n",
      "Std: 164.22856224473418\n",
      "Statistics for 'Flow Packets/s' under 'DoS Hulk'\n",
      "Mean: 180836.96619294072\n",
      "Max: 3000000.0\n",
      "Std: 442867.0566047085\n",
      "Statistics for Non-'DoS Hulk'\n",
      "For 'Fwd Packet Length Max'\n",
      "Mean: [230.6553349304018, 408.7205169628433, 14.932255504860146, 311.76727328809375, 235.63481524249423, 94.67981374965763, 18.9382, 5309.333333333333, 1023.1363636363636, 1.0695330836454433, 323.50242326332796, 54.9072708113804, 277.6666666666667, 22.28048780487805]\n",
      "Max: [24820, 23360, 20, 791, 1983, 410, 49, 5792, 1460, 397, 1432, 602, 600, 585]\n",
      "Std: [791.7018215043946, 2271.518192395181, 6.728071781624097, 199.62902808262837, 427.33344976254864, 111.5918255736925, 5.572198007799761, 747.7439847077786, 409.2625534572369, 3.629565217016018, 321.0237363319064, 165.9429831921441, 290.95120907225333, 110.90248523831782]\n",
      "For 'Flow Packets/s'\n",
      "Mean: [inf, inf, inf, 8.766088353706508, 23251.01982243956, 4560.215782197244, inf, 40.804294023333334, 4559.661403568456, inf, 15474.71827859442, 3420.6611078147344, 16077.049383143416, 2321.9606358469073]\n",
      "Max: [inf, inf, inf, 2587.322122, 2000000.0, 1000000.0, inf, 41.28755526, 100000.0, inf, 2000000.0, 100000.0, 47619.04762, 500000.0]\n",
      "Std: [nan, nan, nan, 108.58675607550104, 143601.1144500906, 23513.993162326606, nan, 0.37550953740010956, 21316.909862043114, nan, 38161.28610535299, 12306.471362662376, 18646.474134614415, 26662.582794585607]\n"
     ]
    }
   ],
   "source": [
    "# Assuming dfs is your list of DataFrames, each representing a different label\n",
    "# And labels_per_group contains the mapping index to label name\n",
    "\n",
    "# Confirm the indices for 'DoS Hulk'\n",
    "hulk_index = labels_per_group.index('DoS Hulk')\n",
    "print(f\"'DoS Hulk' is at index {hulk_index} in labels_per_group\")\n",
    "\n",
    "# Extract 'DoS Hulk' DataFrame\n",
    "hulk_df = dfs[hulk_index]\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Fwd Packet Length Max'\n",
    "fwd_pkt_max_hulk = hulk_df[' Fwd Packet Length Max']\n",
    "print(\"Statistics for 'Fwd Packet Length Max' under 'DoS Hulk'\")\n",
    "print(f\"Mean: {fwd_pkt_max_hulk.mean()}\")\n",
    "print(f\"Max: {fwd_pkt_max_hulk.max()}\")\n",
    "print(f\"Std: {fwd_pkt_max_hulk.std()}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Flow Packets/s'\n",
    "flow_pkts_per_s_hulk = hulk_df[' Flow Packets/s']\n",
    "print(\"Statistics for 'Flow Packets/s' under 'DoS Hulk'\")\n",
    "print(f\"Mean: {flow_pkts_per_s_hulk.mean()}\")\n",
    "print(f\"Max: {flow_pkts_per_s_hulk.max()}\")\n",
    "print(f\"Std: {flow_pkts_per_s_hulk.std()}\")\n",
    "\n",
    "# For Non-'DoS Hulk' (assuming dfs[0] is 'BENIGN' and others are various types of attacks)\n",
    "non_hulk_dfs = [df for i, df in enumerate(dfs) if i != hulk_index]\n",
    "non_hulk_fwd_pkt_max = [df[' Fwd Packet Length Max'] for df in non_hulk_dfs]\n",
    "non_hulk_flow_pkts_per_s = [df[' Flow Packets/s'] for df in non_hulk_dfs]\n",
    "\n",
    "# Stats for Non-'DoS Hulk'\n",
    "print(\"Statistics for Non-'DoS Hulk'\")\n",
    "print(\"For 'Fwd Packet Length Max'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_hulk_fwd_pkt_max]}\")\n",
    "print(f\"Max: {[df.max() for df in non_hulk_fwd_pkt_max]}\")\n",
    "print(f\"Std: {[df.std() for df in non_hulk_fwd_pkt_max]}\")\n",
    "\n",
    "print(\"For 'Flow Packets/s'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_hulk_flow_pkts_per_s]}\")\n",
    "print(f\"Max: {[df.max() for df in non_hulk_flow_pkts_per_s]}\")\n",
    "print(f\"Std: {[df.std() for df in non_hulk_flow_pkts_per_s]}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "bb93e47e-df9d-490a-abfa-cd77fc3d0845",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# Assuming fwd_pkt_max_hulk and non_hulk_fwd_pkt_max are your dataframes or series\n",
    "# You can replace them with your actual dataframes or series\n",
    "\n",
    "# Remove NaN and inf values\n",
    "def remove_invalid_entries(df):\n",
    "    return df.replace([np.inf, -np.inf], np.nan).dropna()\n",
    "\n",
    "# Check if dataframe is empty\n",
    "def is_empty(df):\n",
    "    return df.empty\n",
    "\n",
    "# Fwd Packet Length Max\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "if not is_empty(fwd_pkt_max_hulk):\n",
    "    fwd_pkt_max_hulk = remove_invalid_entries(fwd_pkt_max_hulk)\n",
    "    plt.hist(fwd_pkt_max_hulk, bins=50, alpha=0.5, label='DoS Hulk')\n",
    "\n",
    "non_hulk_concatenated = pd.concat(non_hulk_fwd_pkt_max)\n",
    "if not is_empty(non_hulk_concatenated):\n",
    "    non_hulk_concatenated = remove_invalid_entries(non_hulk_concatenated)\n",
    "    plt.hist(non_hulk_concatenated, bins=50, alpha=0.5, label='Non-DoS Hulk')\n",
    "\n",
    "plt.title('Fwd Packet Length Max')\n",
    "plt.xlabel('Fwd Packet Length Max')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "# Flow Packets/s\n",
    "plt.subplot(1, 2, 2)\n",
    "if not is_empty(flow_pkts_per_s_hulk):\n",
    "    flow_pkts_per_s_hulk = remove_invalid_entries(flow_pkts_per_s_hulk)\n",
    "    plt.hist(flow_pkts_per_s_hulk, bins=50, alpha=0.5, label='DoS Hulk')\n",
    "\n",
    "non_hulk_flow_concatenated = pd.concat(non_hulk_flow_pkts_per_s)\n",
    "if not is_empty(non_hulk_flow_concatenated):\n",
    "    non_hulk_flow_concatenated = remove_invalid_entries(non_hulk_flow_concatenated)\n",
    "    plt.hist(non_hulk_flow_concatenated, bins=50, alpha=0.5, label='Non-DoS Hulk')\n",
    "\n",
    "plt.title('Flow Packets/s')\n",
    "plt.xlabel('Flow Packets/s')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a82fc9d-bf6c-4983-b42c-f3e77f88250a",
   "metadata": {},
   "source": [
    "When dealing with network traffic data to identify types of network activity (e.g., benign vs. malicious like 'DoS Hulk'), various machine learning models can be employed. Below are some commonly used models, the rationale for using them, and considerations based on the statistics you provided.\n",
    "\n",
    "Decision Trees (e.g., CART, Random Forest)\n",
    "Rationale:\n",
    "\n",
    "Easy to interpret and understand, making it possible to understand which features are the most important for classification.\n",
    "Can handle both numerical and categorical variables.\n",
    "Resistant to outliers due to the nature of binary splitting.\n",
    "Evaluation:\n",
    "\n",
    "Given the high variability in your data (Std values), decision trees could be a good choice as they are less affected by outliers.\n",
    "Decision trees may automatically \"threshold\" important features, essentially achieving what your heuristic aimed to do but in a more data-driven manner."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dad95355-ab68-44ac-937d-831d08b6ec44",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "DDoS\n",
    "\n",
    "if ['Flow Packets/s'] > threshold and ['Total Fwd Packets'] > threshold:\n",
    "    return 'DDoS'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "9a33a059-c87c-4a2d-b9c7-c1ddb458fa83",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'DDoS' is at index 2 in labels_per_group\n",
      "Statistics for 'Flow Packets/s' under 'DDoS'\n",
      "Mean: inf\n",
      "Max: inf\n",
      "Std: nan\n",
      "Statistics for 'Total Fwd Packets' under 'DDoS'\n",
      "Mean: 4.474087482642333\n",
      "Max: 9\n",
      "Std: 1.9010806216679734\n",
      "Statistics for Non-'DDoS'\n",
      "For 'Flow Packets/s'\n",
      "Mean: [inf, inf, 8.766088353706508, 180836.96619294072, 23251.01982243956, 4560.215782197244, inf, 40.804294023333334, 4559.661403568456, inf, 15474.71827859442, 3420.6611078147344, 16077.049383143416, 2321.9606358469073]\n",
      "Max: [inf, inf, 2587.322122, 3000000.0, 2000000.0, 1000000.0, inf, 41.28755526, 100000.0, inf, 2000000.0, 100000.0, 47619.04762, 500000.0]\n",
      "Std: [nan, nan, 108.58675607550104, 442867.0566047085, 143601.1144500906, 23513.993162326606, nan, 0.37550953740010956, 21316.909862043114, nan, 38161.28610535299, 12306.471362662376, 18646.474134614415, 26662.582794585607]\n",
      "For 'Total Fwd Packets'\n",
      "Mean: [10.572043386250778, 3.2132471728594507, 5.878161628624306, 5.295635139609314, 5.723729792147806, 6.3566146261298275, 5.458, 2798.5, 795.0, 1.0162496878901373, 11.18578352180937, 13.188619599578503, 2.75, 8.804878048780488]\n",
      "Max: [217797, 38, 21, 15, 19, 16, 9, 2805, 3813, 18, 33, 203, 5, 208]\n",
      "Std: [807.0759703985048, 4.045387479541639, 3.0050537844109226, 2.5652930142814467, 3.2896818630907294, 5.780024357837925, 3.4996979265524364, 5.0099900199501395, 1325.6580396524732, 0.19949002532200896, 10.131186493746009, 44.21930909128054, 1.864744681524183, 33.30412539494919]\n"
     ]
    }
   ],
   "source": [
    "# Assuming dfs is your list of DataFrames, each representing a different label\n",
    "# And labels_per_group contains the mapping index to label name\n",
    "\n",
    "# Find the index for 'DDoS'\n",
    "ddos_index = labels_per_group.index('DDoS')\n",
    "print(f\"'DDoS' is at index {ddos_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'DDoS' DataFrame\n",
    "ddos_df = dfs[ddos_index]\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Flow Packets/s'\n",
    "flow_pkts_per_s_ddos = ddos_df[' Flow Packets/s']\n",
    "print(\"Statistics for 'Flow Packets/s' under 'DDoS'\")\n",
    "print(f\"Mean: {flow_pkts_per_s_ddos.mean()}\")\n",
    "print(f\"Max: {flow_pkts_per_s_ddos.max()}\")\n",
    "print(f\"Std: {flow_pkts_per_s_ddos.std()}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Total Fwd Packets'\n",
    "total_fwd_pkts_ddos = ddos_df[' Total Fwd Packets']\n",
    "print(\"Statistics for 'Total Fwd Packets' under 'DDoS'\")\n",
    "print(f\"Mean: {total_fwd_pkts_ddos.mean()}\")\n",
    "print(f\"Max: {total_fwd_pkts_ddos.max()}\")\n",
    "print(f\"Std: {total_fwd_pkts_ddos.std()}\")\n",
    "\n",
    "# For Non-'DDoS' (assuming dfs[0] is 'BENIGN' and others are various types of attacks)\n",
    "non_ddos_dfs = [df for i, df in enumerate(dfs) if i != ddos_index]\n",
    "non_ddos_flow_pkts_per_s = [df[' Flow Packets/s'] for df in non_ddos_dfs]\n",
    "non_ddos_total_fwd_pkts = [df[' Total Fwd Packets'] for df in non_ddos_dfs]\n",
    "\n",
    "# Stats for Non-'DDoS'\n",
    "print(\"Statistics for Non-'DDoS'\")\n",
    "print(\"For 'Flow Packets/s'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_ddos_flow_pkts_per_s]}\")\n",
    "print(f\"Max: {[df.max() for df in non_ddos_flow_pkts_per_s]}\")\n",
    "print(f\"Std: {[df.std() for df in non_ddos_flow_pkts_per_s]}\")\n",
    "\n",
    "print(\"For 'Total Fwd Packets'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_ddos_total_fwd_pkts]}\")\n",
    "print(f\"Max: {[df.max() for df in non_ddos_total_fwd_pkts]}\")\n",
    "print(f\"Std: {[df.std() for df in non_ddos_total_fwd_pkts]}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ee5e80e4-1076-449f-8da0-e26a2d77da58",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1gAAAGoCAYAAABbkkSYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAA7uUlEQVR4nO3deZgsZXn38e9PQFkVlaOyI0pI0CDikWhwQREDiICJC8QFUYMkGqPRKC4xmNUkbjEkEhRFQAE3kCgqvCqgCYoHZBUQRITDQTlIWFwRvd8/qobTZ+iZ6ZlT3T1z5vu5rrmm66nt7qer66m76qnqVBWSJEmSpDV3n3EHIEmSJElrCxMsSZIkSeqICZYkSZIkdcQES5IkSZI6YoIlSZIkSR0xwZIkSZKkjphgSZIkSVJHTLDmgSRHJ/nrjpa1TZKfJFmnHT47ySu6WHa7vC8kOaSr5c1ivX+f5JYkP+xoeZXkkV0sa1QWaMzPSXJDu00+toPlXZfk50lO6CK+YUiyXftZrTvF+COTnDjK9SZ5efsZLLhtSGvONmag9XbaxnSl3ec9Y4Tr6/TzHGB9I31/0+lyG+jZH/8kyWFdxDcMSV6a5OvTjB/K9jDdepMc17bzy7te7yiZYA1ZzwHhnUluS/K/SQ5Pck/dV9XhVfV3Ay5r2h1RVV1fVRtX1a87iP1eB4JVtU9VfXRNlz3LOLYGXg/sVFUP6zN+jyRnt6+rp/zsJL9od3ATf08cUczHJbmrXeetSc5K8tujWHefWGZ1UJ1ki5l2bO22cWRv3U/hXcCr223y24PGMINnV9WLe2KZ/Jmv1hi0MQ5tRz2fDhB6tXWxB0BVHVtVG483Ig2Dbcyam0sbk+Tynnbl15PamrdMsZ5pT77MIe7edmbi7wVdLHuK9R2Z5Ffteia2tZG0qX1iGXi/O7EvnGi3pphm2m1gDWxaVce061itvezXNvf7TnSl6+2vS73teFW9FNhnfNF0wwRrNJ5dVZsA2wLvBN4EHNv1Subjl6Yj2wI/rqqb5zDvxMH9xN95XQc3jX9pD2q3Am4GjhvhutfEvsAXO1rWtsDlc5kx7Rny+Wot/r5p4bGNWTOzbmOq6lET7QrwNVZva/5xaJHe279MauNOGfL6Tmnf8xLg68BnkmTI6xyFOR9nLITvxUKIcW1jgjVCVXV7VZ0OvAA4JMmj4Z6zUH/fvt4syefas0O3Jvlakvuk6RK1DfDf7dmjN/acjXh5kuuBr0xxhuIRSc5PcnuSzyZ5ULuue53ZnzgrlGRv4C3AC9r1XdyOv+cKQRvX25L8IMnNSY5P8oB23EQchyS5Ps1l97dOVTdJHtDOv7Jd3tva5T8DOAvYoo3juA4+ioHW3Y77QZLHta9f1L6nndrhVyQ5bablV9XPgI8DE5/3v6XpNndHkguSPLknlnWSvCXJ99Kckb6gPbM2OeYntct4Wjv8siRXJPm/JF9Ksm1bfm47y8UTZzen2sZ6Fr8vcEY7/5uS3NjGclWSPQes0/sl+QmwTrvu77Xlv9NuQ7elOQO8f888xyX5QJIzkvwUeNog65qt9vM+NslN7Xv7+6zq7vSIJF9J8uN2m/1Ykk175r2urZNLgJ8mOYlJ38ueVb0syYp2Pa+fJp4npDkTfFuSi9NedRog1nWSvKuN81rgWV3WkxYe25jxtzHTxQxM7I9va9f1xJn2OXOM4dAk/90zfE2ST/QM35Bkl/b1XkmubD+7o4CBkqWq+hXwUeBhwIOTHJFV7dZ3kjxnUkx/kqaNmhi/a5+4fzvJ95Mc1A7vl+SirLpatnNb3m9bXT/JiW093pbkW0keOmB99d0Gkuyfpp26rd0uf6dnnsltwVASmLZOzmq/q1cleX7PuGcl+XaaY4kb0nN1rt93lz7b36rJ8+/tNnBlpmnnM8WxxgCxPjjJ6W2s5wOP6KaG5rGq8m+If8B1wDP6lF8P/Gn7+jjg79vX/wQcDazX/j0ZSL9lAdsBBRwPbARs0FO2bjvN2cCNNAf3GwGfBk5sx+0BLJ8qXuDIiWl7xp8NvKJ9/TLgGmB7YGPgM8AJk2L7YBvXY4BfAr8zRT0dD3wW2KSd97vAy6eKc8C6vyfWPuMKeOQA6z4eeH37+hjgez2f2/HA66ZYfu9nujFNgvW1dvhFwIOBdWm6JPwQWL8d91fApcCONA3dY4AH98YM/AFwA7BbW35g+zn8TrvMtwH/2++9DrCNrQfc0tbFju16tuj5TB8xy8+gt57Xa+N8C3Bf4OnAncCOPXV2O7A7zcmf9Qf9Pk33mU/efoDTgP+i+T48BDgfeGU77pHAXsD9aM7Qngu8b9L6LwK2BjaY4Xt5UruO3wVW0ud7BWwJ/Jgmqb1Pu+4fA0sGiPVw4Mo2lgcBX6Xnuz/T5+Hf2vE31XcC25jJ9dF5G7MGMa/bM+8g+5y++7zez3VS+fbAbTT7lM2BHwA39oz7v3bcZsAdwHPbbeF1wN1M3W7e83m18f4rcEM7/Dxgi3a5LwB+CmzeM+5G4PE07dojgW173x+wK802u19bvitNz4/fozlRd0g77f2m2FZfCfw3sGE7/eOA+8/iM1xtGwB+q30Pe7V188b2c71vz/ovoqctmLS8e33Wfaa51/54Uh1vRNMGH0rTtu9K0z4/qifm323rfGfgR8CBg3532+le2n7mr2vf5wto2uEH9dm2D2SKY40BYj0Z+EQ73aPb7eHrg34eC/HPK1jjs4LmoGiyX9HsELetql9V1deq3dqmcWRV/bSqfj7F+BOq6rKq+inw18Dz0033qxcC76mqa6vqJ8CbgYMmncl5R1X9vKouBi6maQRX08byAuDNVXVnVV0HvBt48eRp5+D97dmn25JcOId1nwM8tX39ZJqDk4nhp7bjp/KGJLfR7JA2ptmRUVUnVtWPq+ruqno3TUO1YzvPK4C3VdVV1bi4qn7cs8zn0SR6+1bV+W3ZK4F/qqorqupu4B+BXXrPLE0y3Tb2FODiqroT+HUb205J1quq66rqe9O835k8oa2Hd1bVXVX1FeBzwME903y2qv6nqn5TVb+Y43p6P/Pb2nUA0J7R3Ad4bfuduRl4L3AQQFVdU1VnVdUvq2ol8B5Wfd73LL+qbpjm+zbhHe06LgU+Mul9TngRcEZVndG+57OAZcC+M8UKPJ/mQOyGqrqVZtuUJtjGtIbcxswl5nsMuM+Zzht69ne3tMu8lubk1S7tsr4E3JjmPuCn0pzs+w3NiZ3vVNWnqrki9T6aE37TeX67X72BJok5sF3nJ6tqRbsfOwW4GtitnecVNF0Zv9W2a9dU1Q96lvlk4HTgkKqa2F//CfBfVfXNqvp1Nffl/ZKmHennVzQnLh/ZTn9BVd0xw3uZzguAz7efza9o7ifeAPj9nmkGbQumc+Gk9uqInnH7AddV1Ufa44ULaU5gPBegqs6uqkvbOr+E5qTe5G1npu8uNIns+9r9wSnAVfTvETHdscaUsbbfvz8C3t7GchnN1c+1mgnW+GwJ3Nqn/F9pDsjPTHJtkiP6TDPZDbMY/wOasxSbDRTl9LZol9e77HWB3svyvTvrn9EcYE+2Gc0VjcnL2rKDGF9TVZu2f/fqkjDAus8BnpzkYTRnxU4Bdk+yHfAAmjNYU3lXu96HVdX+E8lJkte3l9hvb3eoD2DV57E1zVWyqbwW+ER70D5hW+DfenbQt9KcJZyq/qbbxu7pHlhV17TrOxK4OcnJSbaYJraZbEFztvM3PWWTP+eZtuVB9H7mm9Ls+CdsS7P939RTX/9Fc3WIJA9p3+eNSe4ATuTe35VBY5z8vetXd9sCz5vUwD6J9gB4uljb5U1ehzTBNmaVYbYxkw0S8z0G3OdM5109+7ve+c6huQrwlPb12TQH370nBlfbh7SJ9kyf9SfadT2kqp5eVRe07+MlPd35bqO5SjFou3Y4zZWQr/aUbQu8ftK+cWv670cBTqBJJE9O0zX7X5KsN8N7mc5qn2Pbbt1A9+3VrpPaq3f2jNsW+L1JdfBCmm6ZJPm9JF9N0+31dpp6nEt7deOkkyzTtVdTHWtMF+sSmu/AomqvTLDGIMnjaTbIez2isj279vqq2h54NvCXPf1hpzrLONPZx957eLahOdNzC83l7w174lqH5osw6HJX0Hypepd9N81l6tm4pY1p8rJunOVy5mLadbdJxs+A1wDntld2fggcRnN5+zfMQpr7rd5Ec/Xhge0O9XZW9Xu/gen7Jj8PODDJa3vKbqDpNrZpz98GVfW//RYwwza2L/D5nmk/XlVPoqmfAv55Nu93khXA1ln9fq/Jn/NM29yauoHmLOhmPXV1/6p6VDv+n9oYdq6q+9NcYZp8T8LkGKeKefL3bsUU8Zww6bPbqKreOUCsN/VZh2Qbc2+jbGOmi7nf+x1knzMXEwnWk9vXE70xehOs1fYhScLqn+VA2isYHwReTdOlfVPgMgZv1w4Htkny3p6yG4B/mLRv3LCqTmrHr1aX7dWXd1TVTjRXmfYDXjLb99Jjtc+xp25G3V6dM6kONq6qP23Hf5zmyt/WVfUAmq6/07VXU8W7Zfv+JkzXXk11rDFdrCtpvgOLqr0ywRqhJPdPsh9NX9QTJ12FmJhmvySPbDf2O2i6aU08DvdHNP2nZ+tFSXZKsiHwt8CnqnnE7neB9dPcKLkeTX/a+/XM9yNgu0kHxL1OAl6X5OFJNqa5XHxKe+l4YG0snwD+Ickm7c76L2nO5A3VgOs+h6bhmGiUzp40PBub0OxoVgLrJnk7cP+e8R8C/i7JDmnsnOTBPeNXAHsCr0nyZ23Z0cCbkzwK7rmZ+3k986y23Uy1jSV5OE3/9ivb6XZM8vQk9wN+AfycVdviXHyT5oDrjUnWS/Mwh2fTfB9GoqpuAs4E3t1+H++T5ibziW4VmwA/obkJeEuae+JmMtX38q+TbNh+LofSXP2c7ETg2Un+IM1DK9ZP82CArQaI9RM028FWSR7I6l1LtAjZxvQ34jZmuphXAr9h9Tqeyz5nEOfQPChog6paTvOkw71putF9u53m88Cjkvxhmi6Mr6G9OjJLG9EcvK+E5iEbtA91an2Ipivj49p27ZFZvQv7nW1sT0kycQXng8Dh7VWaJNmo3Y42acdPbteeluR32yT+DpqEek3aq08Az0qyZ7vtvp7mhFffE5dD8jngt5K8uG0z10vy+Kx62MYmwK1V9YskuwF/PMPy+m1/0PSKeE27/OfR3GN1Rp/5pzvWmDLW9vv3GeDItk3cieaeurWaCdZo/HeSO2ky/LfS9LE+dIppdwD+H80O9zzgP6vq7HbcPwFvS3P59Q2zWP8JNDfD/hBYn2YnSlXdDvwZzc7vRpqD394nPn2y/f/j9Ll/Cfhwu+xzge/THIT/+Szi6vXn7fqvpTnr+vF2+aMw07rPodmRnTvF8Gx8CfgCzYHHD2jqrPey+Xtoduxn0jQSx9L0+75HVV1Pk2S9KckrqupUmitLJ6fpYnIZq/+GxJHAR9vt5vlMvY09i9V3qvej6a5wC8228xCaB1TMSVXdBezfxnYL8J/ASyYSuhF6CU13oe/Q3Oz9KZoueQDvoLk593aag4/PDLC8qb6X59B0xfoyTTeeMyfPWFU3AAfQ1OtKmm3hr1i1b54u1g/SbE8XAxcOGKvWTrYxMxtVGzNlzNU8UfYfgP9p6/gJzG2fM6Oq+i7NZ/y1dvgOmvf+P+0BL1V1C02viHfSPFxnB+B/5rCu79Dc03YeTeLzu73LqapP0rzvj9MkU6cx6f7AqrqN5oES+yT5u6paRnMf1lE0+75raO9jbk3eVh9Gs3+8A7iCZv875wS6qq6iuZr47zTt1bNpfg7hrrkucw4x3Ak8k+a+2xU0369/ZtVJij8D/rb97r+d5thhuuX12/6gOfm5A837/AfgubX6vd8T8095rDFArK+m6b77Q5p9xUdmURUL0sSTgyQtcknOAI6qqn5nruaFJFfRJBinVtVafwasK+0Z5ffSHPzuVM1N8JKkIWiv0F1Fk2D/VVV9cMwhLRhJjqVJ/G+uqkfONP18ZYIlCYA0v+H077VmT0SSJEla1EywJEmSJKkj3oMlSZIkSR3p+8N3C9Vmm21W22233bjDkCTN0gUXXHBLVS2Zecr5zXZIkhamLtuhtSrB2m677Vi2bNm4w5AkzVKSteKHJ22HJGlh6rIdsougJEmSJHXEBEuSJEmSOmKCJUmSJEkdMcGSJEmSpI6YYEmSJElSR0ywJEmSJKkjJliSJEmS1BETLEmSJEnqiAmWJEmSJHXEBEuSJEmSOmKCJUmSJEkdMcGSJEmSpI6YYEmSJElSR0ywJEmSJKkjJliSJEmS1BETLEmSJEnqyLrjDmA++aM/fjHX3/jDvuO22fJhfPrjJ4w4IknSYjNVW2Q7JEkLgwlWj+tv/CE7H/qPfcdd8pG3jDgaSdJiNFVbZDskSQuDXQQlSZIkqSMmWJIkSZLUERMsSZIkSeqICZYkSZIkdcQES5IkSZI6YoIlSZIkSR0xwZIkSZKkjphgSZIkSVJHTLAkSZIkqSMmWJIkSZLUERMsSZIkSerI0BKsJFsn+WqSK5JcnuQv2vIHJTkrydXt/wdOMf/eSa5Kck2SI4YVpyRJkiR1ZZhXsO4GXl9VvwM8AXhVkp2AI4AvV9UOwJfb4dUkWQf4D2AfYCfg4HZeSZIkSZq3hpZgVdVNVXVh+/pO4ApgS+AA4KPtZB8FDuwz+27ANVV1bVXdBZzczidJkiRJ89ZI7sFKsh3wWOCbwEOr6iZokjDgIX1m2RK4oWd4eVvWb9mHJVmWZNnKlSs7jVuSJEmSZmPoCVaSjYFPA6+tqjsGna1PWfWbsKqOqaqlVbV0yZIlcw1TkiRJktbYUBOsJOvRJFcfq6rPtMU/SrJ5O35z4OY+sy4Htu4Z3gpYMcxYJUmSJGlNDfMpggGOBa6oqvf0jDodOKR9fQjw2T6zfwvYIcnDk9wXOKidT5IkSZLmrWFewdodeDHw9CQXtX/7Au8E9kpyNbBXO0ySLZKcAVBVdwOvBr5E83CMT1TV5UOMVZK0SCX5cJKbk1zWZ9wbklSSzcYRmyRp4Vl3WAuuqq/T/14qgD37TL8C2Ldn+AzgjOFEJ0nSPY4DjgKO7y1MsjXNicDrxxCTJGmBGslTBCVJmq+q6lzg1j6j3gu8kSkesiRJUj8mWJIkTZJkf+DGqrp4gGn9uRBJ0j1MsCRJ6pFkQ+CtwNsHmd6fC5Ek9TLBkiRpdY8AHg5cnOQ6mp8KuTDJw8YalSRpQRjaQy4kSVqIqupS4CETw22StbSqbhlbUJKkBcMrWJKkRS3JScB5wI5Jlid5+bhjkiQtXF7BkiQtalV18AzjtxtRKJKktYBXsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpI+sOa8FJPgzsB9xcVY9uy04Bdmwn2RS4rap26TPvdcCdwK+Bu6tq6bDilCRJkqSuDC3BAo4DjgKOnyioqhdMvE7ybuD2aeZ/WlXdMrToJEmSJKljQ0uwqurcJNv1G5ckwPOBpw9r/ZIkSZI0auO6B+vJwI+q6uopxhdwZpILkhw23YKSHJZkWZJlK1eu7DxQSZIkSRrUuBKsg4GTphm/e1XtCuwDvCrJU6aasKqOqaqlVbV0yZIlXccpSZIkSQMbeYKVZF3gD4FTppqmqla0/28GTgV2G010kiRJkjR347iC9Qzgyqpa3m9kko2SbDLxGngmcNkI45MkSZKkORlagpXkJOA8YMcky5O8vB11EJO6BybZIskZ7eBDga8nuRg4H/h8VX1xWHFKkiRJUleG+RTBg6cof2mfshXAvu3ra4HHDCsuSZIkSRqWcT3kQpIkSZLWOiZYkiRJktQREyxJkiRJ6ogJliRpUUvy4SQ3J7msp+xfk1yZ5JIkpybZdIwhSpIWEBMsSdJidxyw96Sys4BHV9XOwHeBN486KEnSwmSCJUla1KrqXODWSWVnVtXd7eA3gK1GHpgkaUEywZIkaXovA74w1cgkhyVZlmTZypUrRxiWJGk+MsGSJGkKSd4K3A18bKppquqYqlpaVUuXLFkyuuAkSfPS0H5oWJKkhSzJIcB+wJ5VVeOOR5K0MJhgSZI0SZK9gTcBT62qn407HknSwmEXQUnSopbkJOA8YMcky5O8HDgK2AQ4K8lFSY4ea5CSpAXDK1iSpEWtqg7uU3zsyAORJK0VvIIlSZIkSR0xwZIkSZKkjphgSZIkSVJHTLAkSZIkqSMmWJIkSZLUERMsSZIkSeqICZYkSZIkdcQES5IkSZI6YoIlSZIkSR0xwZIkSZKkjphgSZIkSVJHTLAkSZIkqSMmWJIkSZLUERMsSZIkSeqICZYkSZIkdcQES5IkSZI6YoIlSZIkSR0xwZIkSZKkjphgSZIkSVJHTLAkSZIkqSNDS7CSfDjJzUku6yk7MsmNSS5q//adYt69k1yV5JokRwwrRkmSJEnq0jCvYB0H7N2n/L1VtUv7d8bkkUnWAf4D2AfYCTg4yU5DjFOSJEmSOjG0BKuqzgVuncOsuwHXVNW1VXUXcDJwQKfBSZIkSdIQjOMerFcnuaTtQvjAPuO3BG7oGV7elkmSJEnSvDbqBOsDwCOAXYCbgHf3mSZ9ymqqBSY5LMmyJMtWrlzZSZCSJEmSNBcjTbCq6kdV9euq+g3wQZrugJMtB7buGd4KWDHNMo+pqqVVtXTJkiXdBixJkiRJszDSBCvJ5j2DzwEu6zPZt4Adkjw8yX2Bg4DTRxGfJEmSJK2JdYe14CQnAXsAmyVZDvwNsEeSXWi6/F0HvLKddgvgQ1W1b1XdneTVwJeAdYAPV9Xlw4pTkiRJkroytASrqg7uU3zsFNOuAPbtGT4DuNcj3CVJkiRpPhvHUwQlSZIkaa1kgiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkrSoJflwkpuTXNZT9qAkZyW5uv3/wHHGKElaOEywJEmL3XHA3pPKjgC+XFU7AF9uhyVJmpEJliRpUauqc4FbJxUfAHy0ff1R4MBRxiRJWrhMsCRJureHVtVNAO3/h0w1YZLDkixLsmzlypUjC1CSND+ZYEmStAaq6piqWlpVS5csWTLucCRJY2aCJUnSvf0oyeYA7f+bxxyPJGmBMMGSJOneTgcOaV8fAnx2jLFIkhYQEyxJ0qKW5CTgPGDHJMuTvBx4J7BXkquBvdphSZJmtO64A5AkaZyq6uApRu050kAkSWsFr2BJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSODC3BSvLhJDcnuayn7F+TXJnkkiSnJtl0inmvS3JpkouSLBtWjJIkSZLUpWFewToO2HtS2VnAo6tqZ+C7wJunmf9pVbVLVS0dUnySJEmS1KmhJVhVdS5w66SyM6vq7nbwG8BWw1q/JEmSJI3aOO/BehnwhSnGFXBmkguSHDbdQpIclmRZkmUrV67sPEhJkiRJGtRACVaSR3e50iRvBe4GPjbFJLtX1a7APsCrkjxlqmVV1TFVtbSqli5ZsqTLMCVJC0zX7ZUkSbM16BWso5Ocn+TPpnowxaCSHALsB7ywqqrfNFW1ov1/M3AqsNuarFOStGh01l5JkjQXAyVYVfUk4IXA1sCyJB9PstdsV5Zkb+BNwP5V9bMpptkoySYTr4FnApf1m1aSpF5dtVeSJM3VwPdgVdXVwNtoEqSnAu9vH7n+h/2mT3IScB6wY5LlSV4OHAVsApzVPoL96HbaLZKc0c76UODrSS4Gzgc+X1VfnOP7kyQtMrNtryRJ6tK6g0yUZGfgUOBZNI9af3ZVXZhkC5ok6jOT56mqg/ss6th+y2+7BO7bvr4WeMxA0UuS1GMu7ZUkSV0aKMGiufL0QeAtVfXzicKqWpHkbUOJTJKk2bO9kiSN1aAJ1r7Az6vq1wBJ7gOsX1U/q6oThhadJEmzY3slSRqrQe/B+n/ABj3DG7ZlkiTNJ7ZXkqSxGjTBWr+qfjIx0L7ecDghSZI0Z7ZXkqSxGjTB+mmSXScGkjwO+Pk000uSNA62V5KksRr0HqzXAp9MsqId3hx4wVAikiRp7l6L7ZUkaYwGSrCq6ltJfhvYEQhwZVX9aqiRSZI0S7ZXkqRxG/QKFsDjge3aeR6bhKo6fihRSZI0d7ZXkqSxGfSHhk8AHgFcBPy6LS7ABkuSNG/YXkmSxm3QK1hLgZ2qqoYZjCRJa8j2SpI0VoM+RfAy4GHDDESSpA7YXkmSxmrQK1ibAd9Jcj7wy4nCqtp/KFFJkjQ3tleSpLEaNME6cphBSJLUkSO7XFiS1wGvoLmP61Lg0Kr6RZfrkCStXQbqIlhV5wDXAeu1r78FXDjEuCRJmrUu26skWwKvAZZW1aOBdYCDOgpVkrSWGijBSvInwKeA/2qLtgROG1JMkiTNyRDaq3WBDZKsC2wIrJhheknSIjfoQy5eBewO3AFQVVcDDxlWUJIkzVFn7VVV3Qi8C7geuAm4varOnDxdksOSLEuybOXKlXMOXJK0dhg0wfplVd01MdCeyfMRuJKk+aaz9irJA4EDgIcDWwAbJXnR5Omq6piqWlpVS5csWTLHsCVJa4tBE6xzkryFppvEXsAngf8eXliSJM1Jl+3VM4DvV9XKqvoV8Bng9zuKU5K0lho0wToCWEnzBKVXAmcAbxtWUJIkzVGX7dX1wBOSbJgkwJ7AFZ1EKUlaaw30mPaq+g3wwfZPkqR5qcv2qqq+meRTNE8hvBv4NnDMmi5XkrR2GyjBSvJ9+vRhr6rtO49IkqQ56rq9qqq/Af5mTeOSJC0eg/7Q8NKe1+sDzwMe1H04kiStEdsrSdJYDfpDwz/u+buxqt4HPH24oUmSNDu2V5KkcRu0i+CuPYP3oTlDuMlQIpIkaY5sryRJ4zZoF8F397y+G7gOeH7n0UiStGZsryRJYzXoUwSfNuxAJElaU7ZXkqRxG7SL4F9ON76q3tNNOJIkzZ3tlSRp3GbzFMHHA6e3w88GzgVuGEZQkiTNke2VJGmsBk2wNgN2rao7AZIcCXyyql4xrMAkSZoD2ytJ0lgN9Jh2YBvgrp7hu4DtOo9GkqQ1Y3slSRqrQa9gnQCcn+RUoIDnAMcPLSpJkubG9kqSNFaDPkXwH5J8AXhyW3RoVX17eGFJkjR7tleSpHEbtIsgwIbAHVX1b8DyJA+fbuIkH05yc5LLesoelOSsJFe3/x84xbx7J7kqyTVJjphFjJIkzaq9kiSpSwMlWEn+BngT8Oa2aD3gxBlmOw7Ye1LZEcCXq2oH4Mvt8OR1rQP8B7APsBNwcJKdBolTkrS4zbG9kiSpM4NewXoOsD/wU4CqWgFsMt0MVXUucOuk4gOAj7avPwoc2GfW3YBrquraqroLOLmdT5Kkmcy6vZIkqUuDJlh3VVXR3DBMko3muL6HVtVNAO3/h/SZZktW/72S5W1ZX0kOS7IsybKVK1fOMSxJ0lqiq/ZKkqQ5GTTB+kSS/wI2TfInwP8DPjikmNKnrKaauKqOqaqlVbV0yZIlQwpJkrRAjLK9kiTpXmZ8imCSAKcAvw3cAewIvL2qzprD+n6UZPOquinJ5sDNfaZZDmzdM7wVsGIO65IkLSIdt1eSJM3JjAlWVVWS06rqccCaNlKnA4cA72z/f7bPNN8Cdmif+nQjcBDwx2u4XknSWq7j9kqSpDkZtIvgN5I8fjYLTnIScB6wY5LlSV5Ok1jtleRqYK92mCRbJDkDoKruBl4NfAm4AvhEVV0+m3VLkhatWbdXkiR1aaAfGgaeBhye5DqaJzOF5mThzlPNUFUHTzFqzz7TrgD27Rk+AzhjwNgkSZow6/ZKkqQuTZtgJdmmqq6n+U0qSZLmJdsrSdJ8MdMVrNOAXavqB0k+XVV/NIKYJEmardOwvZIkzQMz3YPV+8j07YcZiCRJa8D2SpI0L8yUYNUUryVJmk9sryRJ88JMXQQfk+QOmjODG7SvYdVNw/cfanSSJA3G9kqSNC9Mm2BV1TqjCkSSpLmyvZIkzReD/g6WJEmSJGkGJliSJEmS1BETLEmSJEnqiAmWJEmSJHXEBEuSJEmSOmKCJUmSJEkdMcGSJEmSpI6YYEmSJElSR0ywJEmSJKkjJliSJEmS1BETLEmSJEnqiAmWJElTSLJpkk8luTLJFUmeOO6YJEnz27rjDkCSpHns34AvVtVzk9wX2HDcAUmS5jcTLEmS+khyf+ApwEsBquou4K5xxiRJmv/sIihJUn/bAyuBjyT5dpIPJdlo8kRJDkuyLMmylStXjj5KSdK8YoIlSVJ/6wK7Ah+oqscCPwWOmDxRVR1TVUuraumSJUtGHaMkaZ4xwZIkqb/lwPKq+mY7/CmahEuSpCmZYEmS1EdV/RC4IcmObdGewHfGGJIkaQHwIReSJE3tz4GPtU8QvBY4dMzxSJLmORMsSZKmUFUXAUvHHYckaeGwi6AkSZIkdcQES5IkSZI6YoIlSZIkSR0xwZIkSZKkjphgSZIkSVJHTLAkSZIkqSMjT7CS7Jjkop6/O5K8dtI0eyS5vWeat486TkmSJEmarZH/DlZVXQXsApBkHeBG4NQ+k36tqvYbYWiSJEmStEbG3UVwT+B7VfWDMcchSZIkSWts3AnWQcBJU4x7YpKLk3whyaOmWkCSw5IsS7Js5cqVw4lSkiRJkgYwtgQryX2B/YFP9hl9IbBtVT0G+HfgtKmWU1XHVNXSqlq6ZMmSocQqSZIkSYMY5xWsfYALq+pHk0dU1R1V9ZP29RnAekk2G3WAkiRJkjQb40ywDmaK7oFJHpYk7evdaOL88QhjkyRJkqRZG/lTBAGSbAjsBbyyp+xwgKo6Gngu8KdJ7gZ+DhxUVTWOWCVJkiRpUGNJsKrqZ8CDJ5Ud3fP6KOCoUcclSZIkSWti3E8RlCRJkqS1hgmWJEmSJHXEBEuSJEmSOmKCJUmSJEkdMcGSJEmSpI6YYEmSJElSR0ywJEmSJKkjJliSJEmS1BETLEmSJEnqiAmWJEmSJHXEBEuSJEmSOmKCJUmSJEkdMcGSJEmSpI6YYEmSJElSR0ywJEmSJKkjJliSJEmS1BETLEmSJEnqiAmWJEmSJHXEBEuSJEmSOmKCJUmSJEkdMcGSJEmSpI6YYEmSJElSR0ywJEmaRpJ1knw7yefGHYskaf4zwZIkaXp/AVwx7iAkSQuDCZYkSVNIshXwLOBD445FkrQwmGBJkjS19wFvBH4z5jgkSQuECZYkSX0k2Q+4uaoumGG6w5IsS7Js5cqVI4pOkjRfmWBJktTf7sD+Sa4DTgaenuTEyRNV1TFVtbSqli5ZsmTUMUqS5hkTLEmS+qiqN1fVVlW1HXAQ8JWqetGYw5IkzXMmWJIkSZLUkXXHHYAkSfNdVZ0NnD3mMCRJC4BXsCRJkiSpI2NJsJJcl+TSJBclWdZnfJK8P8k1SS5Jsus44pQkSZKk2RhnF8GnVdUtU4zbB9ih/fs94APtf0mSJEmat+ZrF8EDgOOr8Q1g0ySbjzsoSZIkSZrOuBKsAs5MckGSw/qM3xK4oWd4eVt2L/7AoyRJkqT5YlwJ1u5VtStNV8BXJXnKpPHpM0/1W5A/8ChJkiRpvhhLglVVK9r/NwOnArtNmmQ5sHXP8FbAitFEJ0mSJElzM/IEK8lGSTaZeA08E7hs0mSnAy9pnyb4BOD2qrppxKFKkiRJ0qyM4ymCDwVOTTKx/o9X1ReTHA5QVUcDZwD7AtcAPwMOHUOckiRJkjQrI0+wqupa4DF9yo/ueV3Aq0YZlyRJkiStqfn6mHZJkiRJWnBMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkvpIsnWSrya5IsnlSf5i3DFJkua/dccdgCRJ89TdwOur6sIkmwAXJDmrqr4z7sAkSfOXV7AkSeqjqm6qqgvb13cCVwBbjjcqSdJ8Z4IlSdIMkmwHPBb45phDkSTNcyZYkiRNI8nGwKeB11bVHX3GH5ZkWZJlK1euHH2AkqR5ZeQJ1iA3DSfZI8ntSS5q/94+6jglSUqyHk1y9bGq+ky/aarqmKpaWlVLlyxZMtoAJUnzzjgecjHoTcNfq6r9xhCfJEkkCXAscEVVvWfc8UiSFoaRX8HypmFJ0gKxO/Bi4Ok9PSr2HXdQkqT5bayPaZ/hpuEnJrkYWAG8oaouH2VskqTFraq+DmTccUiSFpaxJVgz3DR8IbBtVf2kPVt4GrDDFMs5DDgMYJttthlewJIkSZI0g7E8RXCmm4ar6o6q+kn7+gxgvSSb9VuWNxdLkiRJmi/G8RTBGW8aTvKwdjqS7EYT549HF6UkSZIkzd44ughO3DR8aZKL2rK3ANsAVNXRwHOBP01yN/Bz4KCqqjHEKkmSJEkDG3mCNchNw1V1FHDUaCKSJEmSpG6M5R4sSZIkSVobmWBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6si64w5AkiRpFP7oj1/M9Tf+8F7l22z5MD798RNmNc9M8813c6mLUZrv8S02o/weTLeuH664kYdtseXQY1hTJliSJGlRuP7GH7Lzof94r/JLPvKWWc8z03zz3VzqYpTme3yLzSi/B9Ot66o3P59nLoDtwi6CkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjriQy4kSTNaW5+kJklS10ywJEkzWlufpCZJUtfsIihJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpIyZYkiRJktQREyxJkiRJ6ogJliRJkiR1xARLkiRJkjpigiVJkiRJHTHBkiRJkqSOmGBJkiRJUkdMsCRJkiSpI2NJsJLsneSqJNckOaLP+CR5fzv+kiS7jiNOSdLiNlN7JUnSZCNPsJKsA/wHsA+wE3Bwkp0mTbYPsEP7dxjwgZEGKUla9AZsryRJWs04rmDtBlxTVddW1V3AycABk6Y5ADi+Gt8ANk2y+agDlSQtaoO0V5IkrSZVNdoVJs8F9q6qV7TDLwZ+r6pe3TPN54B3VtXX2+EvA2+qqmV9lncYzVUugB2Bq9YgvM2AW9Zg/rWJdbGKdbGKdbGKdbFKF3WxbVUt6SKYrgzSXrXli7kdMt7hMt7hMt7hWmjx7lhVm3SxoHW7WMgspU/Z5CxvkGmawqpjgGPWNCiAJMuqamkXy1rorItVrItVrItVrItV1uK6GKgtWsztkPEOl/EOl/EO10KMt6tljaOL4HJg657hrYAVc5hGkqRhsi2SJM3aOBKsbwE7JHl4kvsCBwGnT5rmdOAl7dMEnwDcXlU3jTpQSdKiNkh7JUnSakbeRbCq7k7yauBLwDrAh6vq8iSHt+OPBs4A9gWuAX4GHDqi8Drp4rGWsC5WsS5WsS5WsS5WWSvrYqr2asirXWh1abzDZbzDZbzDtWjjHflDLiRJkiRpbTWWHxqWJEmSpLWRCZYkSZIkdcQEq5Vk7yRXJbkmyRHjjqcLSbZO8tUkVyS5PMlftOUPSnJWkqvb/w/smefNbR1cleQPesofl+TSdtz7k6Qtv1+SU9rybybZbuRvdBaSrJPk2+1vrS3aukiyaZJPJbmy3T6euIjr4nXt9+OyJCclWX+x1EWSDye5OcllPWUjee9JDmnXcXWSQ0b0lue1zNN2KMl17ed7UdrHGM9lOxlifEPdjkcU75FJbmzr+KIk+86HeDOC44gRxTtf63f9JOcnubiN9x1t+Xyt36ninZf1265naMd906qqRf9Hc/Py94DtgfsCFwM7jTuuDt7X5sCu7etNgO8COwH/AhzRlh8B/HP7eqf2vd8PeHhbJ+u0484HnkjzuzBfAPZpy/8MOLp9fRBwyrjf9wx18pfAx4HPtcOLsi6AjwKvaF/fF9h0MdYFsCXwfWCDdvgTwEsXS10ATwF2BS7rKRv6ewceBFzb/n9g+/qB466PMX8W87YdAq4DNptUNuvtZIjxDXU7HlG8RwJv6DPtWONlBMcRI4p3vtZvgI3b1+sB3wSeMI/rd6p452X9tusZ2nHfdH9ewWrsBlxTVddW1V3AycABY45pjVXVTVV1Yfv6TuAKmgPKA2gOsGn/H9i+PgA4uap+WVXfp3mK425JNgfuX1XnVbOlHT9pnollfQrYc1hnIdZUkq2AZwEf6iledHWR5P40DfyxAFV1V1XdxiKsi9a6wAZJ1gU2pPmdo0VRF1V1LnDrpOJRvPc/AM6qqlur6v+As4C9u35/C8xCa4dmtZ0MM5ARbMejiHcqY413RMcRo4h3KuOOt6rqJ+3geu1fMX/rd6p4pzLWeEdw3DclE6zGlsANPcPLmf4LueCk6ZrzWJqzDQ+t9nfF2v8PaSebqh62bF9PLl9tnqq6G7gdePBQ3sSaex/wRuA3PWWLsS62B1YCH2kvm38oyUYswrqoqhuBdwHXAzfR/ObemSzCuugxive+1u9z52A+10kBZya5IMlhbdlst5NR63I7HpVXJ7kkTRfCiW5L8ybeIR5HjCJemKf123Zhuwi4mebE07yu3ynihflZv+9juMd9UzLBavQ7mzxdRr6gJNkY+DTw2qq6Y7pJ+5TVNOXTzTOvJNkPuLmqLhh0lj5la0Vd0Fyx2RX4QFU9FvgpzWXyqay1ddE2AgfQdAfYAtgoyYumm6VP2VpRFwPo8r2vLXXSpflcJ7tX1a7APsCrkjxlmmnn8/uA+btNfgB4BLALzcmed7fl8yLeIR9HdK5PvPO2fqvq11W1C7AVzRWTR08z+XyNd97V74iO+6ZkgtVYDmzdM7wVTTehBS/JejQ7mY9V1Wfa4h+1lzxp/9/clk9VD8vb15PLV5un7WL1AAbv+jBKuwP7J7mOpuvN05OcyOKsi+XA8p6zTp+iSbgWY108A/h+Va2sql8BnwF+n8VZFxNG8d7X2n3uGpi3dVJVK9r/NwOn0nT5m+12MmpdbsdDV1U/ag9cfwN8kFXdKsce7wiOI4Ye73yu3wnVdNU/m6a79Lyt337xztP6HcVx35RMsBrfAnZI8vAk96W5Gfv0Mce0xtp7HY4Frqiq9/SMOh04pH19CPDZnvKD0jz56+HADsD57SXUO5M8oV3mSybNM7Gs5wJfafuozitV9eaq2qqqtqP5fL9SVS9icdbFD4EbkuzYFu0JfIdFWBc0XQOfkGTD9j3sSdNnfzHWxYRRvPcvAc9M8sD2KuIz27LFbF62Q0k2SrLJxGuaz+oyZrmdjDbqe+LoajseuokDvtZzaOp47PGO6Dhi6PHO4/pdkmTT9vUGNCf9rmT+1m/feOdj/Y7ouG/aAPxrjnX2pXnazPeAt447no7e05NoLmNeAlzU/u1Lcw/El4Gr2/8P6pnnrW0dXEXPU1KApTRfmO8BRwFpy9cHPklzM+D5wPbjft8D1MserHqazKKsC5rL+MvabeM0mie5Lda6eAdNg3YZcALNE4QWRV0AJ9F05/gVzVm6l4/qvQMva8uvAQ4dd13Mhz/mYTtEc8/mxe3f5RNxzWU7GWKMQ92ORxTvCcClNPvk04HN50O8jOA4YkTxztf63Rn4dhvXZcDb5/r9GnO887J+e9a1B0M47pvub6IRlCRJkiStIbsISpIkSVJHTLAkSZIkqSMmWJIkSZLUERMsSZIkSeqICZYkSZIkdcQES5pGkl8nuajnb7skeyT5XIfr2C7Jz9vlfyfJ0Ulm9d1sl3HZzFPeM/2BSXYaYLr1kgz6K+iSpFlI8uCe9uWHSW7sGb7vpGlfm2TDAZZ5dpKlU5Rf1bP8584h3pcmOWqK8pU97difdLXsaaYfqD6kcVh33AFI89zPq2qX3oIk2w1hPd+rql2SrAt8BTgQ+MwQ1jPhQOBzND8wPJ0nAf87xDgkadGqqh/T/C4hSY4EflJV75pi8tcCJwI/W4NVvrCqlq3B/NM5papeneQhwOVJTq+qHw1pXdBNfUhD4RUsaQ0keVCS05JckuQbSXZuyy9NsmkaP07ykrb8hCTPmGp5VXU3TULzyCR/kuRbSS5O8umJM3VJHprk1Lb84iS/Pymm7ZN8O8njkzwiyReTXJDka0l+u51+f+Bf27ONj0jymvas4yVJTu5Z3N7AF5JslOTz7fouS/KCbmtSkgSQZM92H35pkg8nuV+S1wBbAF9N8tV2ug8kWZbk8iTvmOO63tgumyTvTfKVnhhObF8fmuS7Sc4Bdp9pmVV1M80Psm47VYxt+/S/bZtyfpJNJsX1rCTnJdksyTPb1xcm+WSSjSfXR5J1khzXtk+XJnndXOpD6ooJljS9DXq6U5zaZ/w7gG9X1c7AW4Dj2/L/oWmIHgVcCzy5LX8C8I2pVtYmUXvS/CL6Z6rq8VX1GOAK4OXtZO8HzmnLdwUu75l/R+DTwKFV9S3gGODPq+pxwBuA/6yq/6X5pfW/qqpdqup7wBHAY9v3cXhPSE8DzqZJtFZU1WOq6tHAF6etNUnSXKwPHAe8oKp+l6an0Z9W1fuBFcDTqupp7bRvraqlwM7AUydO8M3gYz1t2oOBc1nVPi0FNk6yHk3vha8l2Zymndsd2AsYpGv59sD2wDX9YkzT9fEU4C/aduwZwM975n8OTZu0b1v0NuAZVbUrsAz4yz71sQuwZVU9uq23jwxQF9LQ2EVQmt69ughO8iTgjwCq6itp+tM/APga8BTgB8AHgMOSbAncWlU/6bOcRyS5CCjgs1X1hSRPTfL3wKbAxsCX2mmfDrykXeevgduTPBBYAnwW+KOqujzJxsDvA59MMrGe+03xPi6haXhPA04DSLJFG+/PklwKvCvJPwOfq6qvTVMnkqS5WQf4flV9tx3+KPAq4H19pn1+ksNojuU2p0l+Lplh+at1EUxyB/C49grSL4ELaRKtJwOvAX4POLuqVrbTnwL81hTLfkGSJ7XLeWVV3Zrk8D4xFnBTexKQqrqjXTY0J/WWAs+sqjuS7NfO8z/t+PsC5/VZ97XA9kn+Hfg8cOYM9SANlQmWtGbSp6xozgq+CtgGeCvwHOC5NIlXP9/rk8gdBxxYVRcneSmwxwyx3A7cQHOm8XKaK9S3zZAgTngWTUK4P/DXSR4F7EOb1FXVd5M8juaM4j8lObOq/naA5UqSBvfTQSZK8nCaXgmPr6r/S3IczdWvWamqXyW5DjiUpnv6JTRJziNoek78Fk2bNohTqurVA8SYaZZ5Lc3Vr9+iuVoV4KyqOniG9/F/SR4D/AFN2/t84GUDxi11zi6C0po5F3ghQJI9gFuq6o6qugHYDNihqq4Fvk7T0Mzmys8mwE1td40X9pR/GfjTdp3rJLl/W34XzcMrXpLkj9uzgt9P8rx22rQNEMCd7fJJ88TCravqq8AbWXXFbG/gC+00WwA/q6oTgXfRdE2UJHVrfWC7JI9sh18MnNO+vme/DdyfJhm7PclDaU6IzdW5NO3TuTRt1OHARVVVwDeBPdreGesBz5vFcqeK8UpgiySPB0iySZoHPEHT6+MPgePbE33fAHafqI8kGyaZuILW245tBtynqj4N/DW2URozr2BJa+ZI4CNJLqF5ktEhPeO+SdPdA5pG659oEq1B/XW7jB/Q3JM10bD+BXBMkpcDv6ZJtm4CqKqftl0qzkryU5rE7ANJ3gasB5wMXNz+/2B7o/BBwLFt18YA76VpuHaoqivbdf4uzUMxfgP8ql2nJKlbv6C5mvTJNun4FnB0O+4YmocO3VRVT0vybZreCtfS3Pc7V1+j6WlxXtuG/KIto6puSvN0w/No2pkLWdWuTavtfXGvGKvqrjQPSvr3JBvQ3H/1jJ75rkryQuCTwLOBlwInJZno4v424Lv01AfNEwU/klU/cfLmOdSD1Jk0JygkaZW2H/2LqurwGSeWJEnSPUywJEmSJKkj3oMlSZIkSR0xwZIkSZKkjphgSZIkSVJHTLAkSZIkqSMmWJIkSZLUERMsSZIkSerI/weAoMlMDsaiNQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "def plot_histogram(data, title, xlabel, ylabel):\n",
    "    finite_data = data[np.isfinite(data)]\n",
    "    if len(finite_data) > 0:\n",
    "        plt.hist(finite_data, bins=50, edgecolor='black', alpha=0.7)\n",
    "        plt.title(title)\n",
    "        plt.xlabel(xlabel)\n",
    "        plt.ylabel(ylabel)\n",
    "    else:\n",
    "        plt.text(0.5, 0.5, 'No finite data to display', horizontalalignment='center')\n",
    "\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "plot_histogram(ddos_df[' Flow Packets/s'], 'Distribution of \"Flow Packets/s\" for DDoS', 'Flow Packets/s', 'Frequency')\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plot_histogram(ddos_df[' Total Fwd Packets'], 'Distribution of \"Total Fwd Packets\" for DDoS', 'Total Fwd Packets', 'Frequency')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# For Non-'DDoS'\n",
    "for i, df in enumerate(non_ddos_dfs):\n",
    "    label = labels_per_group[i] if i != ddos_index else \"Other\"\n",
    "\n",
    "    plt.figure(figsize=(12, 6))\n",
    "\n",
    "    plt.subplot(1, 2, 1)\n",
    "    plot_histogram(df[' Flow Packets/s'], f'Distribution of \"Flow Packets/s\" for {label}', 'Flow Packets/s', 'Frequency')\n",
    "\n",
    "    plt.subplot(1, 2, 2)\n",
    "    plot_histogram(df[' Total Fwd Packets'], f'Distribution of \"Total Fwd Packets\" for {label}', 'Total Fwd Packets', 'Frequency')\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d60aac4-0da4-4fcd-977a-1e15e3eb275c",
   "metadata": {},
   "source": [
    "Evaluation of Heuristic\n",
    "Given the output statistics, here are some observations:\n",
    "\n",
    "'Flow Packets/s' under 'DDoS': The mean and max values are inf, and the standard deviation is nan. This suggests that there might be some anomalies or extreme values in your data. You should investigate these extreme or infinite values to understand what's causing this. Without addressing these, using this feature as part of a heuristic might not be very reliable.\n",
    "\n",
    "'Total Fwd Packets' under 'DDoS': The mean is approximately 4.47, the maximum is 9, and the standard deviation is approximately 1.90. This suggests that for 'DDoS' attacks, this metric could serve as a distinguishing factor.\n",
    "\n",
    "Non-'DDoS' Statistics: For 'Flow Packets/s', the mean, max, and standard deviation vary widely for non-DDoS traffic. For 'Total Fwd Packets', the mean, max, and standard deviation are also different but less varied compared to 'Flow Packets/s'.\n",
    "\n",
    "The heuristic if ['Flow Packets/s'] > threshold and ['Total Fwd Packets'] > threshold: return 'DDoS' is problematic due to the following reasons:\n",
    "\n",
    "Data Anomalies: You'd need to resolve why 'Flow Packets/s' under 'DDoS' are inf. If there are outliers, they could distort your heuristic.\n",
    "\n",
    "Overlap: Even if we exclude 'DDoS', the values of 'Flow Packets/s' and 'Total Fwd Packets' for non-DDoS labels vary widely and overlap with potential 'DDoS' values, making it hard to choose a \"threshold\" that cleanly separates 'DDoS' from non-'DDoS' traffic.\n",
    "\n",
    "Machine Learning Models\n",
    "Given the variability in your data, machine learning models are a better choice than heuristics for this kind of problem. Here are some suggestions prioritized by likely effectiveness:\n",
    "\n",
    "1.Random Forest Classifier\n",
    "\n",
    "Why: Random Forests are good with high dimensionality and can capture complex relationships in the data. They also offer feature importance scores, helping you understand which features are most relevant.\n",
    "\n",
    "2.Gradient Boosting Machine (e.g., XGBoost)\n",
    "\n",
    "Why: Similar to Random Forests but generally more powerful, they can capture non-linear relationships and are robust to outliers. They are also good for imbalanced classification problems, which might be the case here.\n",
    "\n",
    "3.Support Vector Machine (SVM) with RBF Kernel\n",
    "\n",
    "Why: Effective for high-dimensional spaces and also capable of modeling non-linear decision boundaries. However, they might be computationally expensive for large datasets."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e103ce2-3add-4400-9534-117b840cb9ab",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "PortScan: Scanning multiple ports to find an open port.\n",
    "if ['Destination Port'] > threshold and ['Fwd IAT Max'] < threshold:\n",
    "    return 'PortScan'\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "ed4ac2fd-2cac-427d-bb37-931a3403f643",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'PortScan' is at index 10 in labels_per_group\n",
      "Statistics for ' Destination Port' under 'PortScan'\n",
      "Mean: 8629.93484144819\n",
      "Max: 65389\n",
      "Std: 13475.6892963097\n",
      "Statistics for ' Fwd IAT Max' under 'PortScan'\n",
      "Mean: 76093.77281398252\n",
      "Max: 119000000\n",
      "Std: 2204307.9588743844\n",
      "Statistics for Non-'PortScan'\n",
      "For ' Destination Port'\n",
      "Mean: [9407.82391272463, 17560.41114701131, 81.94824935528665, 80.0, 80.0, 80.0, 80.0, 21.0, 444.0, 444.0, 22.0, 80.0, 80.0, 80.0]\n",
      "Max: [65534, 53938, 64873, 80, 80, 80, 80, 21, 444, 444, 22, 80, 80, 80]\n",
      "Std: [19745.242209782715, 19017.78880711812, 336.9055571454257, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]\n",
      "For ' Fwd IAT Max'\n",
      "Mean: [4325915.816910301, 176186.23909531502, 15634986.690153243, 19463654.40592227, 56942576.161081836, 38896295.66657044, 40636917.30621748, 1585949.2694, 1163159.0, 37363899.5, 1225405.5102315564, 4760417.766069547, 843593.1666666666, 5126719.743902439]\n",
      "Max: [120000000, 10200000, 101000000, 119000000, 118000000, 110000000, 119000000, 3941973, 1996118, 104000000, 6027793, 5996344, 5000673, 5993177]\n",
      "Std: [14598417.403702047, 354153.069922709, 28947603.317146707, 29075973.24649769, 45613611.375557065, 34497227.662367634, 39982077.24501041, 1612330.8659819588, 408065.0063644272, 30906918.423951983, 1231756.693999287, 1761684.3728488693, 1941792.5112434754, 1138630.425099872]\n"
     ]
    }
   ],
   "source": [
    "# Import pandas if not imported\n",
    "import pandas as pd\n",
    "\n",
    "# Assuming dfs is your list of DataFrames, each representing a different label\n",
    "# And labels_per_group contains the mapping index to label name\n",
    "\n",
    "# Find the index for 'PortScan'\n",
    "portscan_index = labels_per_group.index('PortScan')\n",
    "print(f\"'PortScan' is at index {portscan_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'PortScan' DataFrame\n",
    "portscan_df = dfs[portscan_index]\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for ' Destination Port'\n",
    "dest_port_portscan = portscan_df[' Destination Port']\n",
    "print(\"Statistics for ' Destination Port' under 'PortScan'\")\n",
    "print(f\"Mean: {dest_port_portscan.mean()}\")\n",
    "print(f\"Max: {dest_port_portscan.max()}\")\n",
    "print(f\"Std: {dest_port_portscan.std()}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for ' Fwd IAT Max'\n",
    "fwd_iat_max_portscan = portscan_df[' Fwd IAT Max']\n",
    "print(\"Statistics for ' Fwd IAT Max' under 'PortScan'\")\n",
    "print(f\"Mean: {fwd_iat_max_portscan.mean()}\")\n",
    "print(f\"Max: {fwd_iat_max_portscan.max()}\")\n",
    "print(f\"Std: {fwd_iat_max_portscan.std()}\")\n",
    "\n",
    "# For Non-'PortScan'\n",
    "non_portscan_dfs = [df for i, df in enumerate(dfs) if i != portscan_index]\n",
    "non_portscan_dest_port = [df[' Destination Port'] for df in non_portscan_dfs]\n",
    "non_portscan_fwd_iat_max = [df[' Fwd IAT Max'] for df in non_portscan_dfs]\n",
    "\n",
    "# Stats for Non-'PortScan'\n",
    "print(\"Statistics for Non-'PortScan'\")\n",
    "print(\"For ' Destination Port'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_portscan_dest_port]}\")\n",
    "print(f\"Max: {[df.max() for df in non_portscan_dest_port]}\")\n",
    "print(f\"Std: {[df.std() for df in non_portscan_dest_port]}\")\n",
    "\n",
    "print(\"For ' Fwd IAT Max'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_portscan_fwd_iat_max]}\")\n",
    "print(f\"Max: {[df.max() for df in non_portscan_fwd_iat_max]}\")\n",
    "print(f\"Std: {[df.std() for df in non_portscan_fwd_iat_max]}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "2de6ab1e-e6d3-4728-97e0-ec491caf784f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Import matplotlib if not imported\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "# Set up the figure and axis\n",
    "fig, axs = plt.subplots(2, 2, figsize=(14, 10))\n",
    "fig.suptitle('Data Distribution for \"PortScan\" and Non-\"PortScan\" Labels')\n",
    "\n",
    "# Plot for ' Destination Port' under 'PortScan'\n",
    "axs[0, 0].hist(dest_port_portscan, bins=50, edgecolor='black', alpha=0.7)\n",
    "axs[0, 0].set_title('Distribution of \"Destination Port\" under \"PortScan\"')\n",
    "axs[0, 0].set_xlabel('Destination Port')\n",
    "axs[0, 0].set_ylabel('Frequency')\n",
    "\n",
    "# Plot for ' Destination Port' under Non-'PortScan'\n",
    "# Concatenate all dataframes into one for this variable\n",
    "all_non_portscan_dest_port = pd.concat(non_portscan_dest_port)\n",
    "axs[0, 1].hist(all_non_portscan_dest_port, bins=50, edgecolor='black', alpha=0.7)\n",
    "axs[0, 1].set_title('Distribution of \"Destination Port\" under Non-\"PortScan\"')\n",
    "axs[0, 1].set_xlabel('Destination Port')\n",
    "axs[0, 1].set_ylabel('Frequency')\n",
    "\n",
    "# Plot for ' Fwd IAT Max' under 'PortScan'\n",
    "axs[1, 0].hist(fwd_iat_max_portscan, bins=50, edgecolor='black', alpha=0.7)\n",
    "axs[1, 0].set_title('Distribution of \"Fwd IAT Max\" under \"PortScan\"')\n",
    "axs[1, 0].set_xlabel('Fwd IAT Max')\n",
    "axs[1, 0].set_ylabel('Frequency')\n",
    "\n",
    "# Plot for ' Fwd IAT Max' under Non-'PortScan'\n",
    "# Concatenate all dataframes into one for this variable\n",
    "all_non_portscan_fwd_iat_max = pd.concat(non_portscan_fwd_iat_max)\n",
    "axs[1, 1].hist(all_non_portscan_fwd_iat_max, bins=50, edgecolor='black', alpha=0.7)\n",
    "axs[1, 1].set_title('Distribution of \"Fwd IAT Max\" under Non-\"PortScan\"')\n",
    "axs[1, 1].set_xlabel('Fwd IAT Max')\n",
    "axs[1, 1].set_ylabel('Frequency')\n",
    "\n",
    "# Display the plots\n",
    "plt.tight_layout(rect=[0, 0.03, 1, 0.95])\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7d9e7ab8-218e-4224-a3cc-cea52540387a",
   "metadata": {},
   "source": [
    "Heuristic Evaluation:\n",
    "For the heuristic: if ['Destination Port'] > threshold and ['Fwd IAT Max'] < threshold: return 'PortScan'\n",
    "\n",
    "Based on the given statistics:\n",
    "\n",
    "Mean of ['Destination Port'] under 'PortScan' is approximately 8629.93, while the means for Non-'PortScan' vary widely.\n",
    "\n",
    "Max of ['Destination Port'] under 'PortScan' is 65389, while the max for Non-'PortScan' also varies but mostly not exceeding 65534.\n",
    "\n",
    "Standard deviation for ['Destination Port'] under 'PortScan' is 13475.69.\n",
    "\n",
    "Mean of ['Fwd IAT Max'] under 'PortScan' is approximately 76093.77, while the means for Non-'PortScan' generally are much higher.\n",
    "\n",
    "Max of ['Fwd IAT Max'] under 'PortScan' is 119000000, which is very high but comparable with Non-'PortScan'.\n",
    "\n",
    "Standard deviation for ['Fwd IAT Max'] under 'PortScan' is 2204307.96.\n",
    "\n",
    "Given these statistics, a reasonable heuristic threshold might be somewhere near the mean or median for each feature. However, due to the wide range and high standard deviation, especially for ['Fwd IAT Max'], the heuristic could result in a high number of false positives or negatives.\n",
    "\n",
    "Machine Learning Models:\n",
    "Random Forest Classifier\n",
    "\n",
    "Argument: Random Forests are generally good at handling imbalanced datasets and irrelevant input variables. Given that our data has some features with very high variance and potentially high cardinality, Random Forests can give a better classification performance.\n",
    "Evaluation: High variance in ['Fwd IAT Max'] and presence of outliers can be easily tackled.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e32ee50-8178-4884-bb6c-2e4aa22b48b7",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'DoS GoldenEye':\n",
    "if ['Fwd Packets/s'] > threshold and ['Bwd Packet Length Max'] < threshold:\n",
    "    return 'DoS GoldenEye'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "df4bf985-6043-4da1-87c1-af5cac43ae77",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'DoS GoldenEye' is at index 3 in labels_per_group\n",
      "Statistics for 'Fwd Packets/s' under 'DoS GoldenEye'\n",
      "Mean: 8.402375795992906\n",
      "Max: 2587.322122\n",
      "Std: 108.61162085405643\n",
      "Statistics for 'Bwd Packet Length Max' under 'DoS GoldenEye'\n",
      "Mean: 4152.516347933374\n",
      "Max: 11632\n",
      "Std: 3426.028850565815\n",
      "Statistics for Non-'DoS GoldenEye'\n",
      "For 'Fwd Packets/s'\n",
      "Mean: [58336.44929953641, 21909.755234135733, 110.52567466844931, 180434.16712970092, 11520.914323777937, 2267.239656366967, 114975.01260700004, 23.463037056666668, 4552.569026617091, 31338.01166634469, 7824.46186999524, 1713.5656554089821, 8038.566174310584, 1834.2213117334634]\n",
      "Max: [3000000.0, 1000000.0, 1500000.0, 3000000.0, 1000000.0, 500000.0, 2000000.0, 23.51222893, 100000.0, 1000000.0, 1000000.0, 50000.0, 23809.52381, 500000.0]\n",
      "Std: [231401.64974840987, 72616.15479380582, 7316.509489861263, 442206.809140501, 71789.52527484816, 11759.058063592704, 303831.14302315755, 0.0408124061593136, 21318.485434522125, 63781.258494291775, 20868.570963942635, 6157.231458959035, 9323.198053848027, 25193.607828152708]\n",
      "For 'Bwd Packet Length Max'\n",
      "Mean: [397.1838083797205, 50.95799676898223, 4617.122594723269, 4017.27503172939, 106.0918013856813, 16.742262393864696, 16.83, 14480.0, 17.318181818181817, 8.767051186017477, 492.46742057081315, 80.97576396206533, 1025.1666666666667, 98.73414634146341]\n",
      "Max: [19530, 256, 11595, 11595, 3525, 5792, 34, 17376, 267, 5792, 976, 5330, 4149, 7926]\n",
      "Std: [814.167347501526, 59.12300193654473, 3958.508270235089, 3123.3406081183034, 426.30027106598527, 136.04031883102635, 17.000850148776568, 1831.5912207695253, 55.795021422922815, 118.93467812005389, 488.0452588359023, 302.8608640919007, 1329.0414750579616, 620.7488104122655]\n"
     ]
    }
   ],
   "source": [
    "# Import pandas if not imported\n",
    "import pandas as pd\n",
    "\n",
    "# Assuming dfs is your list of DataFrames, each representing a different label\n",
    "# And labels_per_group contains the mapping index to label name\n",
    "\n",
    "# Find the index for 'DoS GoldenEye'\n",
    "goldeneye_index = labels_per_group.index('DoS GoldenEye')\n",
    "print(f\"'DoS GoldenEye' is at index {goldeneye_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'DoS GoldenEye' DataFrame\n",
    "goldeneye_df = dfs[goldeneye_index]\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Fwd Packets/s'\n",
    "fwd_pkts_per_s_goldeneye = goldeneye_df['Fwd Packets/s']\n",
    "print(\"Statistics for 'Fwd Packets/s' under 'DoS GoldenEye'\")\n",
    "print(f\"Mean: {fwd_pkts_per_s_goldeneye.mean()}\")\n",
    "print(f\"Max: {fwd_pkts_per_s_goldeneye.max()}\")\n",
    "print(f\"Std: {fwd_pkts_per_s_goldeneye.std()}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Bwd Packet Length Max'\n",
    "bwd_pkt_len_max_goldeneye = goldeneye_df['Bwd Packet Length Max']\n",
    "print(\"Statistics for 'Bwd Packet Length Max' under 'DoS GoldenEye'\")\n",
    "print(f\"Mean: {bwd_pkt_len_max_goldeneye.mean()}\")\n",
    "print(f\"Max: {bwd_pkt_len_max_goldeneye.max()}\")\n",
    "print(f\"Std: {bwd_pkt_len_max_goldeneye.std()}\")\n",
    "\n",
    "# For Non-'DoS GoldenEye'\n",
    "non_goldeneye_dfs = [df for i, df in enumerate(dfs) if i != goldeneye_index]\n",
    "non_goldeneye_fwd_pkts_per_s = [df['Fwd Packets/s'] for df in non_goldeneye_dfs]\n",
    "non_goldeneye_bwd_pkt_len_max = [df['Bwd Packet Length Max'] for df in non_goldeneye_dfs]\n",
    "\n",
    "# Stats for Non-'DoS GoldenEye'\n",
    "print(\"Statistics for Non-'DoS GoldenEye'\")\n",
    "print(\"For 'Fwd Packets/s'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_goldeneye_fwd_pkts_per_s]}\")\n",
    "print(f\"Max: {[df.max() for df in non_goldeneye_fwd_pkts_per_s]}\")\n",
    "print(f\"Std: {[df.std() for df in non_goldeneye_fwd_pkts_per_s]}\")\n",
    "\n",
    "print(\"For 'Bwd Packet Length Max'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_goldeneye_bwd_pkt_len_max]}\")\n",
    "print(f\"Max: {[df.max() for df in non_goldeneye_bwd_pkt_len_max]}\")\n",
    "print(f\"Std: {[df.std() for df in non_goldeneye_bwd_pkt_len_max]}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "be6b6779-2ce4-45d3-9a30-518dce6c92b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Generate summary statistics for 'Fwd Packets/s' and 'Bwd Packet Length Max'\n",
    "goldeneye_stats = {\n",
    "    'Fwd Packets/s': {\n",
    "        'Mean': fwd_pkts_per_s_goldeneye.mean(),\n",
    "        'Max': fwd_pkts_per_s_goldeneye.max(),\n",
    "        'Std': fwd_pkts_per_s_goldeneye.std()\n",
    "    },\n",
    "    'Bwd Packet Length Max': {\n",
    "        'Mean': bwd_pkt_len_max_goldeneye.mean(),\n",
    "        'Max': bwd_pkt_len_max_goldeneye.max(),\n",
    "        'Std': bwd_pkt_len_max_goldeneye.std()\n",
    "    }\n",
    "}\n",
    "\n",
    "# Generate summary statistics for Non-'DoS GoldenEye' cases\n",
    "non_goldeneye_stats = {\n",
    "    'Fwd Packets/s': {\n",
    "        'Mean': [df.mean() for df in non_goldeneye_fwd_pkts_per_s],\n",
    "        'Max': [df.max() for df in non_goldeneye_fwd_pkts_per_s],\n",
    "        'Std': [df.std() for df in non_goldeneye_fwd_pkts_per_s]\n",
    "    },\n",
    "    'Bwd Packet Length Max': {\n",
    "        'Mean': [df.mean() for df in non_goldeneye_bwd_pkt_len_max],\n",
    "        'Max': [df.max() for df in non_goldeneye_bwd_pkt_len_max],\n",
    "        'Std': [df.std() for df in non_goldeneye_bwd_pkt_len_max]\n",
    "    }\n",
    "}\n",
    "\n",
    "# Plot summary statistics for 'Fwd Packets/s'\n",
    "plt.figure(figsize=(10, 6))\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.title(\"Summary for 'Fwd Packets/s'\")\n",
    "plt.bar(goldeneye_stats['Fwd Packets/s'].keys(), goldeneye_stats['Fwd Packets/s'].values(), alpha=0.7, label='DoS GoldenEye')\n",
    "plt.ylabel('Value')\n",
    "plt.legend()\n",
    "\n",
    "# Plot summary statistics for 'Bwd Packet Length Max'\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.title(\"Summary for 'Bwd Packet Length Max'\")\n",
    "plt.bar(goldeneye_stats['Bwd Packet Length Max'].keys(), goldeneye_stats['Bwd Packet Length Max'].values(), alpha=0.7, label='DoS GoldenEye')\n",
    "plt.ylabel('Value')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Plot histograms for 'Fwd Packets/s' and 'Bwd Packet Length Max' for both 'DoS GoldenEye' and Non-'DoS GoldenEye'\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "# Histogram for 'Fwd Packets/s'\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(fwd_pkts_per_s_goldeneye, alpha=0.5, label='DoS GoldenEye', bins=20)\n",
    "plt.hist([item for sublist in non_goldeneye_fwd_pkts_per_s for item in sublist], alpha=0.5, label='Non-DoS GoldenEye', bins=20)\n",
    "plt.title('Histogram of Fwd Packets/s')\n",
    "plt.xlabel('Fwd Packets/s')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "# Histogram for 'Bwd Packet Length Max'\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.hist(bwd_pkt_len_max_goldeneye, alpha=0.5, label='DoS GoldenEye', bins=20)\n",
    "plt.hist([item for sublist in non_goldeneye_bwd_pkt_len_max for item in sublist], alpha=0.5, label='Non-DoS GoldenEye', bins=20)\n",
    "plt.title('Histogram of Bwd Packet Length Max')\n",
    "plt.xlabel('Bwd Packet Length Max')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f479abe8-25bd-4429-8c9f-e2aea5db6624",
   "metadata": {},
   "source": [
    "Evaluation of Heuristic:\n",
    "Heuristic:\n",
    "if ['Fwd Packets/s'] > threshold and ['Bwd Packet Length Max'] < threshold\n",
    "then, return 'DoS GoldenEye'.\n",
    "\n",
    "Observations:\n",
    "'Fwd Packets/s' for 'DoS GoldenEye' has a Mean of 8.4, Max of 2587.32, and Std of 108.6.\n",
    "'Bwd Packet Length Max' for 'DoS GoldenEye' has a Mean of 4152.5, Max of 11632, and Std of 3426.0.\n",
    "The other classes show significantly higher Mean and Max values for 'Fwd Packets/s', with values often going up to millions.\n",
    "For 'Bwd Packet Length Max', the other classes generally have a lower Mean and Max value compared to 'DoS GoldenEye', although there are exceptions.\n",
    "Evaluation:\n",
    "The heuristic is likely to have many false negatives for 'DoS GoldenEye' because its 'Fwd Packets/s' Mean is relatively low (8.4) compared to other classes (which go up to millions).\n",
    "It may have false positives if the threshold for 'Bwd Packet Length Max' is not well-chosen, as other classes also have instances where 'Bwd Packet Length Max' can be high.\n",
    "The standard deviation is relatively high for both features in 'DoS GoldenEye', implying variability in the data.\n",
    "Machine Learning Models:\n",
    "1. Random Forest Classifier\n",
    "Why: Handles feature interactions well, and does not require scaling of features. Random forests can capture complex patterns in the data without the need for extensive feature engineering.\n",
    "Evaluation: Given the high variance in features across classes, a Random Forest could do well in segregating 'DoS GoldenEye' from others.\n",
    "2. Gradient Boosting Machines (e.g., XGBoost)\n",
    "Why: Effective for imbalanced datasets and can automatically handle missing values. It is also less prone to overfitting.\n",
    "Evaluation: Given the large standard deviations and potential outliers, boosting algorithms may give good performance.\n",
    "3. Support Vector Machines (SVM)\n",
    "Why: Effective in high-dimensional spaces and also when the number of dimensions is greater than the number of samples. Kernels can be used for non-linear separations.\n",
    "Evaluation: May work well if a linear separation is not possible between classes. However, feature scaling is essential, and it might be computationally expensive.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb5aeb70-9612-4809-8e45-c37a00227487",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'FTP-Patator':\n",
    "if ['Destination Port'] == 21 and ['Fwd Packet Length Mean'] > threshold:\n",
    "    return 'FTP-Patator'\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "04de6adb-2726-42e1-863a-49bb63475405",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'FTP-Patator' is at index 7 in labels_per_group\n",
      "Statistics for 'Fwd Packet Length Mean' under 'FTP-Patator' with port 21\n",
      "Mean: 9.359669444457\n",
      "Max: 15.0\n",
      "Std: 2.4926717706871613\n",
      "Statistics for Non-'FTP-Patator'\n",
      "Mean: [66.43078926458519, 116.21841861694669, 7.40588951368109, 59.20787826049969, 44.579436380856166, 158.2734136135335, 63.51428676409833, 5.1522990821666665, 301.98209193181816, 1.0080582605077655, 48.10517343858373, 17.219615082086406, 62.18333333333333, 8.535335342682925]\n",
      "Max: [4672.0, 5675.444444, 10.0, 398.0625, 317.25, 1983.0, 239.0, 7.443968594, 920.75, 147.3, 174.1818182, 216.5073892, 134.25, 241.3054187]\n",
      "Std: [204.2573453007653, 616.4774016077347, 1.2361810644729843, 56.251254207751614, 39.534729901827326, 407.3398286028672, 78.70837439201327, 1.1508994041850165, 187.87939733601297, 1.340298636782255, 47.792422381220995, 53.72818435672608, 65.3384347884617, 43.07720889365892]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Find the index for 'FTP-Patator'\n",
    "ftp_patator_index = labels_per_group.index('FTP-Patator')\n",
    "print(f\"'FTP-Patator' is at index {ftp_patator_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'FTP-Patator' DataFrame\n",
    "ftp_patator_df = dfs[ftp_patator_index]\n",
    "\n",
    "# Calculate statistics for 'Destination Port' and 'Fwd Packet Length Mean'\n",
    "dest_port_ftp = ftp_patator_df[' Destination Port']\n",
    "fwd_pkt_len_mean_ftp = ftp_patator_df[' Fwd Packet Length Mean']\n",
    "\n",
    "# Filter to only those rows where 'Destination Port' is 21\n",
    "ftp_patator_filtered_df = ftp_patator_df[ftp_patator_df[' Destination Port'] == 21]\n",
    "\n",
    "# Calculate statistics for 'Fwd Packet Length Mean' after filtering\n",
    "fwd_pkt_len_mean_ftp_filtered = ftp_patator_filtered_df[' Fwd Packet Length Mean']\n",
    "print(\"Statistics for 'Fwd Packet Length Mean' under 'FTP-Patator' with port 21\")\n",
    "print(f\"Mean: {fwd_pkt_len_mean_ftp_filtered.mean()}\")\n",
    "print(f\"Max: {fwd_pkt_len_mean_ftp_filtered.max()}\")\n",
    "print(f\"Std: {fwd_pkt_len_mean_ftp_filtered.std()}\")\n",
    "\n",
    "# For Non-'FTP-Patator'\n",
    "non_ftp_patator_dfs = [df for i, df in enumerate(dfs) if i != ftp_patator_index]\n",
    "non_ftp_patator_fwd_pkt_len_mean = [df[' Fwd Packet Length Mean'] for df in non_ftp_patator_dfs]\n",
    "\n",
    "# Stats for Non-'FTP-Patator'\n",
    "print(\"Statistics for Non-'FTP-Patator'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_ftp_patator_fwd_pkt_len_mean]}\")\n",
    "print(f\"Max: {[df.max() for df in non_ftp_patator_fwd_pkt_len_mean]}\")\n",
    "print(f\"Std: {[df.std() for df in non_ftp_patator_fwd_pkt_len_mean]}\")\n",
    "\n",
    "# Visualization using Matplotlib\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "# Histogram for 'Fwd Packet Length Mean' for 'FTP-Patator'\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(fwd_pkt_len_mean_ftp_filtered, bins=30, color='blue', alpha=0.7, label='FTP-Patator')\n",
    "plt.title('Fwd Packet Length Mean for FTP-Patator')\n",
    "plt.xlabel('Fwd Packet Length Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "# Histogram for 'Fwd Packet Length Mean' for Non-'FTP-Patator'\n",
    "plt.subplot(1, 2, 2)\n",
    "for df in non_ftp_patator_fwd_pkt_len_mean:\n",
    "    plt.hist(df, bins=30, alpha=0.5, label='Non-FTP-Patator')\n",
    "plt.title('Fwd Packet Length Mean for Non-FTP-Patator')\n",
    "plt.xlabel('Fwd Packet Length Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f72d40c3-0372-45d6-a8b0-85028e65a5bf",
   "metadata": {},
   "source": [
    "### Evaluation of Heuristic\n",
    "\n",
    "**Heuristic**: if `['Destination Port'] == 21` and `['Fwd Packet Length Mean'] > threshold`: return 'FTP-Patator'\n",
    "\n",
    "Based on the statistics provided:\n",
    "\n",
    "- **Mean of 'Fwd Packet Length Mean' for 'FTP-Patator' with port 21**: 9.36\n",
    "- **Max**: 15.0\n",
    "- **Standard Deviation**: 2.49\n",
    "- **Mean for Non-'FTP-Patator' ranges**: between 1.00 and 301.98\n",
    "- **Max for Non-'FTP-Patator' ranges**: between 7.44 and 5675.44\n",
    "- **Standard Deviation for Non-'FTP-Patator' ranges**: between 1.15 and 616.47\n",
    "\n",
    "We see that the mean value for 'Fwd Packet Length Mean' in 'FTP-Patator' cases is considerably lower than most of the means in the non-'FTP-Patator' cases. Therefore, setting a threshold should be carefully considered. A lower threshold could potentially lead to a lot of false positives, while a higher one might miss actual 'FTP-Patator' cases.\n",
    "\n",
    "Given these statistics, a threshold around the mean value for 'FTP-Patator' (9.36) could be a starting point, but further validation is needed. Also, the standard deviation for 'FTP-Patator' suggests that the data isn't very dispersed (Std: 2.49), which could make the heuristic fairly reliable for this particular label.\n",
    "\n",
    "### Machine Learning Models for Distinguishing Between the Cases\n",
    "\n",
    "1. **Random Forest Classifier**\n",
    "    - **Why**: Random Forest is robust to outliers and can handle both categorical and numerical features effectively. Given the variety of features and the likely non-linear relationships between them, Random Forest might be an excellent choice.\n",
    "    - **Evaluation based on data**: Since the mean and standard deviation for 'FTP-Patator' and Non-'FTP-Patator' are quite different, Random Forest could likely separate them effectively using feature importance.\n",
    "\n",
    "2. **Gradient Boosting Machines (XGBoost, LightGBM)**\n",
    "    - **Why**: These are effective for imbalanced datasets and can model complex relationships between features.\n",
    "    - **Evaluation based on data**: The large range of Max values and Std in Non-'FTP-Patator' could be captured well with gradient-boosted trees.\n",
    "\n",
    "3. **Support Vector Machines (SVM)**\n",
    "    - **Why**: SVM works well for a clear margin of separation and is effective in high-dimensional spaces.\n",
    "    - **Evaluation based on data**: Given the lower mean and std values for 'FTP-Patator', a hyperplane could potentially separate it effectively from non-'FTP-Patator' categories.\n",
    "  \n",
    "4. **K-Nearest Neighbors (KNN)**\n",
    "    - **Why**: KNN is straightforward and can be highly effective if the classes form distinct clusters.\n",
    "    - **Evaluation based on data**: KNN could be less reliable here because of the overlapping means and std deviations across categories. However, with proper normalization, it might be effective.\n",
    "\n",
    "5. **Logistic Regression**\n",
    "    - **Why**: If the relationship between the labels and features is approximately linear, logistic regression can be a simple yet effective model.\n",
    "    - **Evaluation based on data**: Given the diversity of Std and Max values in Non-'FTP-Patator', the data might not be linearly separable, making Logistic Regression less optimal here.\n",
    "\n",
    "Based on these factors, Random Forest would likely be the most effective, followed by gradient boosting machines, SVM, KNN, and lastly, Logistic Regression."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6d580ae-7ef8-40fa-a4b9-2e9e7cbd5189",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'DoS slowloris':\n",
    "if ['Fwd IAT Mean'] > threshold and ['Fwd IAT Max'] > threshold:\n",
    "    return 'DoS slowloris'\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "f8029228-fb1b-4267-bfba-53d5a16dff3e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'DoS slowloris' is at index 6 in labels_per_group\n",
      "Statistics for 'Fwd IAT Mean' under 'DoS slowloris'\n",
      "Mean: 25124858.235790994\n",
      "Max: 119000000.0\n",
      "Std: 39973270.41135584\n",
      "Statistics for 'Fwd IAT Max' under 'DoS slowloris'\n",
      "Mean: 40636917.30621748\n",
      "Max: 119000000\n",
      "Std: 39982077.24501041\n",
      "Statistics for Non-'DoS slowloris'\n",
      "For 'Fwd IAT Mean'\n",
      "Mean: [1895659.9960036646, 148017.65809574397, 2670881.32377665, 15962772.298253153, 10138945.367458042, 12803522.11382416, 373055.54647714, 42633.170425, 2408829.658339091, 73872.91469998968, 257404.8205301831, 2296949.1532757957, 211540.0902778333, 2508872.4671457075]\n",
      "Max: [120000000.0, 6396441.875, 40700000.0, 119000000.0, 59100000.0, 36700000.0, 946785.5714, 42730.98065, 7484994.357, 119000000.0, 728235.0, 2998568.5, 1251865.0, 2996990.0]\n",
      "Std: [9265739.834163815, 280513.9400597901, 4319184.61893551, 30611518.685626965, 8813465.274023954, 10242335.028955314, 378649.5790705356, 77.7611859934388, 2464492.616471749, 2196212.818226852, 257528.27929489585, 988374.5236727425, 485855.10144306306, 687727.203348014]\n",
      "For 'Fwd IAT Max'\n",
      "Mean: [4325915.816910301, 176186.23909531502, 15634986.690153243, 19463654.40592227, 56942576.161081836, 38896295.66657044, 1585949.2694, 1163159.0, 37363899.5, 76093.77281398252, 1225405.5102315564, 4760417.766069547, 843593.1666666666, 5126719.743902439]\n",
      "Max: [120000000, 10200000, 101000000, 119000000, 118000000, 110000000, 3941973, 1996118, 104000000, 119000000, 6027793, 5996344, 5000673, 5993177]\n",
      "Std: [14598417.403702047, 354153.069922709, 28947603.317146707, 29075973.24649769, 45613611.375557065, 34497227.662367634, 1612330.8659819588, 408065.0063644272, 30906918.423951983, 2204307.9588743844, 1231756.693999287, 1761684.3728488693, 1941792.5112434754, 1138630.425099872]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Find the index for 'DoS slowloris'\n",
    "slowloris_index = labels_per_group.index('DoS slowloris')\n",
    "print(f\"'DoS slowloris' is at index {slowloris_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'DoS slowloris' DataFrame\n",
    "slowloris_df = dfs[slowloris_index]\n",
    "\n",
    "# Calculate the statistics for 'Fwd IAT Mean' and 'Fwd IAT Max'\n",
    "fwd_iat_mean_slowloris = slowloris_df[' Fwd IAT Mean']\n",
    "fwd_iat_max_slowloris = slowloris_df[' Fwd IAT Max']\n",
    "\n",
    "print(\"Statistics for 'Fwd IAT Mean' under 'DoS slowloris'\")\n",
    "print(f\"Mean: {fwd_iat_mean_slowloris.mean()}\")\n",
    "print(f\"Max: {fwd_iat_mean_slowloris.max()}\")\n",
    "print(f\"Std: {fwd_iat_mean_slowloris.std()}\")\n",
    "\n",
    "print(\"Statistics for 'Fwd IAT Max' under 'DoS slowloris'\")\n",
    "print(f\"Mean: {fwd_iat_max_slowloris.mean()}\")\n",
    "print(f\"Max: {fwd_iat_max_slowloris.max()}\")\n",
    "print(f\"Std: {fwd_iat_max_slowloris.std()}\")\n",
    "\n",
    "# For Non-'DoS slowloris' \n",
    "non_slowloris_dfs = [df for i, df in enumerate(dfs) if i != slowloris_index]\n",
    "non_slowloris_fwd_iat_mean = [df[' Fwd IAT Mean'] for df in non_slowloris_dfs]\n",
    "non_slowloris_fwd_iat_max = [df[' Fwd IAT Max'] for df in non_slowloris_dfs]\n",
    "\n",
    "# Stats for Non-'DoS slowloris'\n",
    "print(\"Statistics for Non-'DoS slowloris'\")\n",
    "print(\"For 'Fwd IAT Mean'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_slowloris_fwd_iat_mean]}\")\n",
    "print(f\"Max: {[df.max() for df in non_slowloris_fwd_iat_mean]}\")\n",
    "print(f\"Std: {[df.std() for df in non_slowloris_fwd_iat_mean]}\")\n",
    "\n",
    "print(\"For 'Fwd IAT Max'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_slowloris_fwd_iat_max]}\")\n",
    "print(f\"Max: {[df.max() for df in non_slowloris_fwd_iat_max]}\")\n",
    "print(f\"Std: {[df.std() for df in non_slowloris_fwd_iat_max]}\")\n",
    "\n",
    "# Visualization using matplotlib\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(fwd_iat_mean_slowloris, bins=30, color='blue', alpha=0.7, label='DoS slowloris')\n",
    "plt.title('Distribution of Fwd IAT Mean for DoS slowloris')\n",
    "plt.xlabel('Fwd IAT Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.hist(fwd_iat_max_slowloris, bins=30, color='red', alpha=0.7, label='DoS slowloris')\n",
    "plt.title('Distribution of Fwd IAT Max for DoS slowloris')\n",
    "plt.xlabel('Fwd IAT Max')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39e9bf93-416b-4cfa-a8c9-bde97daa63fc",
   "metadata": {},
   "source": [
    "### Evaluating the Heuristic\n",
    "\n",
    "To evaluate the heuristic, we'll have to set some thresholds for the 'Fwd IAT Mean' and 'Fwd IAT Max' variables for detecting 'DoS slowloris' attacks.\n",
    "\n",
    "From the provided statistics:\n",
    "\n",
    "- 'Fwd IAT Mean' in 'DoS slowloris': \n",
    "  - Mean: ~25,124,858\n",
    "  - Max: 119,000,000\n",
    "  - Std: ~39,973,270\n",
    "  \n",
    "- 'Fwd IAT Max' in 'DoS slowloris':\n",
    "  - Mean: ~40,636,917\n",
    "  - Max: 119,000,000\n",
    "  - Std: ~39,982,077\n",
    "\n",
    "For the non-'DoS slowloris' cases, the statistics for the same variables are much lower on average, but there are high maximums and variances, so we'll need to be cautious about setting the thresholds too low, as it could lead to false positives.\n",
    "\n",
    "Given these statistics, a simple heuristic might be to set a threshold somewhat above the mean for 'Fwd IAT Mean' and 'Fwd IAT Max' for the 'DoS slowloris' cases, yet sufficiently high to avoid overlaps with most non-'DoS slowloris' traffic. However, a more robust approach would involve using machine learning models.\n",
    "\n",
    "### Machine Learning Models\n",
    "\n",
    "Here are some machine learning models that could be effective for this binary classification problem:\n",
    "\n",
    "#### 1. Random Forest Classifier\n",
    "- **Why:** Random forests are generally good at handling high-dimensional data and can capture complex interactions between features.\n",
    "- **Evaluation:** Given the high variance in the data (as indicated by the standard deviations), a Random Forest model would likely capture this complexity effectively.\n",
    "\n",
    "#### 2. Support Vector Machines (SVM)\n",
    "- **Why:** SVMs are effective for binary classification problems and can work well when the data is not linearly separable, which might be the case here.\n",
    "- **Evaluation:** The variance is high for the 'DoS slowloris' category. An SVM with a radial basis function (RBF) kernel could potentially create a decision boundary that effectively separates the two classes.\n",
    "\n",
    "#### 3. Gradient Boosting Machines (e.g., XGBoost, LightGBM)\n",
    "- **Why:** Like Random Forests, these ensemble methods can handle complex feature interactions but often provide better performance.\n",
    "- **Evaluation:** These models are known for high accuracy and can be tuned to handle overfitting, which could be a concern given the high variance in the data.\n",
    "\n",
    "#### 4. Neural Networks\n",
    "- **Why:** Neural networks are capable of capturing complex relationships in the data.\n",
    "- **Evaluation:** Given the high dimensionality and potential complexity of the interactions between features, neural networks could perform well. However, they might be overkill for this problem and are prone to overfitting if not carefully tuned.\n",
    "\n",
    "#### 5. Logistic Regression\n",
    "- **Why:** This is a simple and interpretable model that works well for binary classification problems.\n",
    "- **Evaluation:** It may struggle if the relationship between the features and the target variable is complex or non-linear, but it's worth trying as a baseline model.\n",
    "\n",
    "To choose the best model, cross-validation techniques could be employed to compare performance on unseen data, followed by hyperparameter tuning for the best-performing models. The imbalanced nature of attack vs non-attack classes may also require techniques like SMOTE for oversampling the minority class or adjusting class weights."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ebc4fa9a-78f4-46d6-b869-baa9bd88c25e",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'DoS Slowhttptest':\n",
    "if ['Destination Port'] == 80 and ['Fwd IAT Mean'] > threshold:\n",
    "    return 'DoS Slowhttptest'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "7a3ea385-66ce-4501-860e-0a93a0cff4ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'DoS Slowhttptest' is at index 5 in labels_per_group\n",
      "Statistics for 'Destination Port' under 'DoS Slowhttptest'\n",
      "Unique Ports: [80]\n",
      "Most Common Port: 80\n",
      "Statistics for 'Fwd IAT Mean' under 'DoS Slowhttptest'\n",
      "Mean: 12803522.11382416\n",
      "Max: 36700000.0\n",
      "Std: 10242335.028955314\n",
      "Statistics for Non-'DoS Slowhttptest'\n",
      "For 'Destination Port'\n",
      "Unique Ports: [array([   53,   123,    80, ..., 34049, 26699,  4306]), array([53720,  8080, 52316, 52256, 53513,  2108, 51596, 51774,  1845,\n",
      "       52397,  2937, 52940,  2876, 51631,  4076,  4199, 51845, 52861,\n",
      "        2878, 51745, 53508,  2964, 52917,  4302, 52345, 51666,  4196,\n",
      "       51633, 51779, 52317, 53713,  4306, 52237,  2363,  4299, 51664,\n",
      "       53524, 53031, 52344, 51703,  4295, 51682, 53344, 53721,  3051,\n",
      "        2911, 52869, 53029, 53742, 52857,  2903, 51713, 52928, 52860,\n",
      "        2751, 52324,  2853,  4184,  4087,  1991,  3847,  5070, 52725,\n",
      "        3210,  4212,  2877, 51702, 52093,  1851, 53793,  3046, 53876,\n",
      "       51665, 53894, 51717, 53932, 53447, 52222,  3632, 51695,  3388,\n",
      "        4286, 53723,  4089, 52907, 52708,  3621, 52932, 52707,  3642,\n",
      "       52837,  3382, 53736,  2952, 52921,  2910, 51663, 52242, 53518,\n",
      "        2972,  3205, 51720, 52709,  2953, 53529, 51760, 52175, 51757,\n",
      "       51701, 52312, 51704, 51662, 52732,  3842, 53730, 51706, 52924,\n",
      "       52901, 50054, 52939, 52230, 51733, 53507, 52947, 53026, 52239,\n",
      "       52935, 52342,  3846, 52234,  3395,  2909, 51595,  1993, 53734,\n",
      "       52726, 53548, 51599, 52140,  1847, 51683, 53107, 52330, 52265,\n",
      "        1848, 53547, 51622,  4282,  1846, 52323, 52347,  2920, 52831,\n",
      "        2861,  1859, 52235,  2864,  4073, 51773, 53097, 52180,  2967,\n",
      "        3618,  4179, 52258, 53509, 51598,  2915, 52913, 51726, 51734,\n",
      "       53433, 52325, 52176,  3392, 52713, 51698, 52233,  4300,  4069,\n",
      "       51780, 52243,  1841,  4173, 53028,  4078,  3050, 52169,  4084,\n",
      "       52148,  2353, 52948, 53099, 52198, 53541, 51656,  2756, 51697,\n",
      "        4273, 52716, 53938, 53710, 51721, 51609, 51741, 53722, 52909,\n",
      "        4176,  2995,  4284, 53525, 52311, 52174,  2908, 51737, 52873,\n",
      "       52699, 52178, 51786, 52251, 51632,  2997, 51629, 51799, 52933,\n",
      "        2354, 52729, 52329, 51686,  2872, 53892, 52252,  2982,  3013,\n",
      "       52908, 51712, 53434,  3829,  2912, 52738, 53727,  2844,  3830,\n",
      "       53523,  3652,  3400, 52318,  2994,  2871, 51754, 51613, 53881,\n",
      "       51763, 53725, 53728, 53733, 52868,  4071, 52236, 52337, 53897,\n",
      "       52914, 52930, 53337, 51611,  3653, 52872, 52264, 51604, 53884,\n",
      "        3650,  4201, 53543, 52740,  2956,  3643, 53376, 53893,  2978,\n",
      "       50059, 53709, 53539, 53544, 51586,  3645, 51705,  2857,  4204,\n",
      "       52789, 53888,  4297, 51782, 53593,  2918, 52332,  2925, 51777,\n",
      "       53446,  2959, 51634, 51787, 52922,  2973, 52723, 51653, 52262,\n",
      "        4617,  4278,  2869, 51707, 51687, 53043, 53514, 53889,  2757,\n",
      "       51623, 53520, 52931,  2764,  4182, 53719, 51711, 53035, 51612,\n",
      "       53526, 53901, 51597,  2987, 51727, 53896, 52870, 53024,  2979,\n",
      "        4194,  2451, 52338, 51681, 51583, 51710,  3836,  2870,  3843,\n",
      "        4866, 52339, 51740, 53030, 52942,  2850,  2750, 53440, 51637,\n",
      "        4303,  2989, 51722, 51585,  2914, 51755, 53714, 52349, 51790,\n",
      "       52263,  4285,  3841, 53931,  4086, 52240,  4858,  2968,  2741,\n",
      "       51606, 51781,  2916, 52259, 51646,  3021, 53890, 52737, 51743,\n",
      "       51764, 53885, 51800,  2749,  2922, 51738,  2572, 52255, 52321,\n",
      "        2846, 51592,  3014,  3026, 52179, 51783, 52254,  4088,  4077,\n",
      "       53899,  3206, 51791, 51725, 51739,  3422, 52257, 51831, 53027,\n",
      "        2996,  3207, 51608,  4277, 53935,  2855,  2849, 53900,  2574,\n",
      "        2923,  2894, 53735,  4068, 51801,  2848, 52248, 53032, 51699,\n",
      "       52336, 51772,  3045,  2957,  4294, 51588,  2107, 51789,  2919,\n",
      "       52181, 52206, 51809,  3617,  2954,  2965, 51759]), array([   80, 27636, 64873, 64869]), array([80]), array([80]), array([80]), array([21]), array([444]), array([444]), array([ 3737, 17877,  6059,    30, 58080,  1102,  1999,  9103,  1064,\n",
      "       18040,  1594,     1, 32783, 44176,  2399,   125, 64680,  2608,\n",
      "        3006,  2170,  4224,   222,  1079,   720,  2522,  4444,  3914,\n",
      "         163,   765,  3971,  9929,  3283, 15004,  6112, 10180,  1971,\n",
      "       10004,  1119,  6106, 15660,  3268,  3322,  7000,  1024,  3517,\n",
      "        9091, 50002,  9071,   255,  5850,  5405,  2107,  2366,  3325,\n",
      "        3878,  5440,  9080, 19801, 23502, 55600,  3828,  2021,  8099,\n",
      "       10000, 16012,  8089,  8654,  2394,     4,  3221,   111,  5988,\n",
      "        5004,  1059,  1046,  7200,  3211,  6001,    25,  6699,  1021,\n",
      "        8008,  1022, 17988,  3071,  2393,   106,  9500,  2041,  1048,\n",
      "        5903,  5080,  1067,  1113,  2038, 57797, 19283,  2718,   143,\n",
      "        3546,    82,   993,  2047,   211,  2111, 32776,  3827,  5226,\n",
      "        5003,  9101, 50800, 64623,  1164,   873,  1272,  8383, 10628,\n",
      "        4443,   443, 49999, 22939,  2811, 49400,  2602,  5862, 40193,\n",
      "       56737,  5009,  1174,  1035,  8002,  7201, 49156,  2013,   212,\n",
      "         544,   898,  2022, 14442,  1011, 34573,  2382,  1072,  4449,\n",
      "        1717, 13783,  3389,  9898,  8000,  9593,  6025,  1045,   777,\n",
      "        3351,  2043,  6881,  1147,  1108,  5214,  5718,    42,  6646,\n",
      "        1688,  5190,  7778,   687,  1068,  4006,  6006,   548, 13722,\n",
      "        1213,  1007,  8400, 49159,  3371,  1086,  5222,  9998,  1123,\n",
      "        2605,   711,    83,  5859,   444, 50001,  1166, 33354,  9081,\n",
      "        3077,  4279,  9050,  2710,  1148,  1034,  5000, 49158, 21571,\n",
      "         110,  9535,  1002,  3476,    90,  3390,   161,  2251,  1248,\n",
      "        1216,  2383,  1183,  5298, 10082,  5510, 15002,  7025,  1031,\n",
      "        4343,  2967,    22,  1247,  1163,  1065,  4125,  2869,  1028,\n",
      "        3260,  1914,    17, 10617,  3814, 28201,  9111, 32772,  1271,\n",
      "       14441,  1500, 10012,   555,    21,  2003,  5002, 55055, 50000,\n",
      "        1862, 27355,  2260,  2875,  1029, 35500,  3031, 10025,  5800,\n",
      "        9876,   100, 50389,  6002,  1580,     6,  9943, 30951,  1122,\n",
      "        2500,  9594,  2068,  5900,  9001,  3851,  6689,  7920,  9207,\n",
      "        2030,  2920, 65389, 32782,    33,  5960,   464,  1277,   901,\n",
      "        1236,   427, 56738, 49176, 27000,  1296,  1124,  3030,  1175,\n",
      "        5730, 16113,  1069,  1083,   593,    20,  3003, 50636,  8181,\n",
      "        3261, 13456,  4848, 34572,   683,  5431,  2557,  1217,  8011,\n",
      "        5906, 49161,  1700, 50006,  1322,  1111,  5907,  3300,   512,\n",
      "        6567, 14000,  7007,  6666,  6101,  6009,  1054,   616,  8800,\n",
      "         666,  2401,    88,  1864,  2492,  6510,  1042,  8045,  1132,\n",
      "        1071, 10024,  3128, 10243,  3659,  3372,  6543,   787,  3404,\n",
      "        3306,  2105,  1057,  3945,  2005,    26,  2604,  1900,  5999,\n",
      "        9485,  8888,  4126, 49155,  2065,   541,  6566,   554,  1434,\n",
      "        9502,  1076,  3801,  1040,  7800,  1084,   801,  4242,     7,\n",
      "        5815,  8994,  1998,    79, 32774, 51103,   109, 20031,   981,\n",
      "        5432,  4662,  2196,  1666,  9595,   648,  2020,  8007,  1091,\n",
      "        2008,  1524,  4550,  2800, 32780,  5961,  9666,  2106, 15742,\n",
      "        5877,   617,  1723,  2045,  1095,  1145,  1105,   990,  7496,\n",
      "        3690, 33899,  9110, 62078,  3301,  7001,  3784,  6788,  1117,\n",
      "        2002,  1009,  3766,  8600,  6000,  5950,   691,  5822,  2601,\n",
      "       32778,  1187, 65129,  9090,  2048,  8192,  2725,  3001,  8042,\n",
      "       49163,   543,   843,  1149,  2381,    19,  8031,  1066,  5989,\n",
      "        3369,  3269,   301,  1863,  1098,  6100,  1010,   903, 11967,\n",
      "        8402,   992,  1080,  1687,  9002,  8010,  1092,   587, 32771,\n",
      "        3869,  1801,  1074,   995,  5120,  5666,    13,  9200, 32768,\n",
      "       24444,  9618, 18988,  2126,  2119,  6003,  5952,  9415, 32781,\n",
      "           3,  5102,  6580,  1110,  1718,   306,  2121, 10621,  1027,\n",
      "        3871,  4005, 10009,  2288,  5633,  8443,  2160,  5963,   646,\n",
      "        4111,  5087,  7019,  2222,    80, 19780,  2607,  2135,  1328,\n",
      "         179, 61532,  1974,  1060,  1037,  4321,  5030,  8899, 16001,\n",
      "        5962,  4446,  2717,  5560,   668,  1030, 54328,  9290,  5904,\n",
      "        1805,  6156,  9999,  1782,  1062,  1311,  7435,  1036,  9220,\n",
      "        3580,  1840,  1073, 65000, 60020,  8652, 26214,  1186,  5033,\n",
      "        8085,  1761,  3324,   406, 49160,  4899, 32777, 13782,  1085,\n",
      "        2910,  3880,  6005, 24800,  3995,  9099,   880,  1001,  1049,\n",
      "       10616, 10010,  1839,  1055, 49165, 10002,  5101, 42510,   256,\n",
      "        1287, 51493,   900,   515,  3011,    24,  1096,  6901,  9968,\n",
      "        1494,  2007, 49175,  7004, 11111,  5566,  3905, 44442,  1192,\n",
      "       49152,  2809,  1169,  3367,  3005,  1050,  4000, 63331,  3013,\n",
      "         425, 10001,  5357,  6779,  2099, 14238,  5414,  7911,  8022,\n",
      "        4567,  5500,   513, 10003,  2046, 27356,  8873,  6839,   254,\n",
      "       10215,   912, 32785,  5555,  1039,  5061,  9418,  5051,  6669,\n",
      "        4004,  3493, 41511,  6565,  5679,  3551,  1082,  5100,   340,\n",
      "        8300,  1077,  8292,    84, 12000, 10629, 38292,  1301,  5544,\n",
      "        1094,  1154,  1052,  1309, 57294,  1947,  1334,  7676,  5801,\n",
      "       16000,  1600, 15000,  5987,  6969,   783,  5925,  8081,  1556,\n",
      "        3998, 30718,  5631,   911,    53,  7103,  1259,  3826,  7777,\n",
      "        6792,  2998,  6667,  2010,  5811,  1104, 20221,  6129,  2323,\n",
      "       18101,  2100,  2179,  2968,  6123,  1433, 10626,  3168,   416,\n",
      "        4445,  3800,  1130,  2525,  5825, 16992,  1521,   445,  1417,\n",
      "        9878,  8290,   800,  5200, 55555,  4998,  8100,  2909,   366,\n",
      "        1300,  7002,  9000,  3527,  9011,  9900,   500,  1106,  1234,\n",
      "        3052,  8021, 11110,  8087, 44443,  8084,  1131,  1443, 32769,\n",
      "       20000,   199, 54045,  5902,  9102, 16993,  1090,  5550,  1201,\n",
      "       40911,  8180,  3000, 32784,  2042,   631,  2004,   999,  8500,\n",
      "        5001, 61900, 27715, 19101,  2034,  4003,  3333,   722,  1000,\n",
      "       20005, 52869, 44501,  1025, 10778,  5810, 49157, 32770,  7921,\n",
      "        9944,   667,   749,  2701, 16080,   458,  1043,    23,  8701,\n",
      "        1185,  3703, 20828,  5280,  2190,  5911,  5922,   987,  7100,\n",
      "        1218,  9917, 20222,   625, 30000,  4900,  2040,  1070, 25735,\n",
      "        7741,  5998,   311,  2638,  1755,  1081,  8200,  5959,    70,\n",
      "        7512,  1089, 12265,  8651,   417,  1078,  9003,  1138,  9877,\n",
      "          99,  3007, 50003,  1310,  6692,  1152,  1721, 27352,  6547,\n",
      "        1719, 55056,  5910,  8222, 19350,  9100,  1875,  1063,  1461,\n",
      "        8649,  2035, 32773,  7625,  1352,  1051,  4045,  1023, 32775,\n",
      "        1107,  8254,  2009,  7106,  1114,   259,   714,   407,   465,\n",
      "        1783,  1137,  1503,  2033,  2103,  7938, 16016,  1641,  1233,\n",
      "       50500,   139,   636,   146,   563,  6346,  1041,  1583, 12174,\n",
      "        1061,  5050,   700,     9,   264, 12345,  8333,  4001,  2001,\n",
      "        3809,  3323,  1935,  1099,   389, 49167,    37,   497, 16018,\n",
      "       52822,  7937,    81,  6789,  1151, 31337,  1199,  3920,    85,\n",
      "        8088,  6502, 19315,  1198,  1033,  8083,  1047,  7627,  3370,\n",
      "          89,  2006,  2144,  1244,  7443,  1501,  1026, 34571,  1044,\n",
      "        8080,   514,  1455,  1126,  9040,  1097,  5269,   902,   280,\n",
      "         705,  1088,  7070,   808,  5678, 27353,  8093,    43,  9575,\n",
      "        9010,  3918,  1141,  5915,  5221,  1720,  1658,  8291,  8009,\n",
      "        1812,  1056,  8082, 45100,  1984,  1100,  2049, 19842,  1112,\n",
      "        3689,    49,  6389,  8090,  7999,  4129, 49154,  8193,   144,\n",
      "        1075, 49153, 52848,    32, 32779,  2301,  1972,  5901, 25734,\n",
      "        6668,  8086,  6004,  1032,  3889,   135,  2200, 10566,   726,\n",
      "        1533,  5802, 31038, 48080,  1087,  1058,  1038,   119,  5225,\n",
      "        9009,  1053,  6007, 15003,   545,   481,  2161,  8001,  1093,\n",
      "        1165,  7402,  3986,  4002,  9503,  2702,   524,  5054,  3017,\n",
      "        2191, 52673,  8194, 60443,  1121,   888, 50300,  2000,     0]), array([22]), array([80]), array([80]), array([80])]\n",
      "For 'Fwd IAT Mean'\n",
      "Mean: [1895659.9960036646, 148017.65809574397, 2670881.32377665, 15962772.298253153, 10138945.367458042, 25124858.235790994, 373055.54647714, 42633.170425, 2408829.658339091, 73872.91469998968, 257404.8205301831, 2296949.1532757957, 211540.0902778333, 2508872.4671457075]\n",
      "Max: [120000000.0, 6396441.875, 40700000.0, 119000000.0, 59100000.0, 119000000.0, 946785.5714, 42730.98065, 7484994.357, 119000000.0, 728235.0, 2998568.5, 1251865.0, 2996990.0]\n",
      "Std: [9265739.834163815, 280513.9400597901, 4319184.61893551, 30611518.685626965, 8813465.274023954, 39973270.41135584, 378649.5790705356, 77.7611859934388, 2464492.616471749, 2196212.818226852, 257528.27929489585, 988374.5236727425, 485855.10144306306, 687727.203348014]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Find the index for 'DoS Slowhttptest'\n",
    "slowhttptest_index = labels_per_group.index('DoS Slowhttptest')\n",
    "print(f\"'DoS Slowhttptest' is at index {slowhttptest_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'DoS Slowhttptest' DataFrame\n",
    "slowhttptest_df = dfs[slowhttptest_index]\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Destination Port'\n",
    "destination_port_slowhttptest = slowhttptest_df[' Destination Port']\n",
    "print(\"Statistics for 'Destination Port' under 'DoS Slowhttptest'\")\n",
    "print(f\"Unique Ports: {destination_port_slowhttptest.unique()}\")\n",
    "print(f\"Most Common Port: {destination_port_slowhttptest.mode()[0]}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Fwd IAT Mean'\n",
    "fwd_iat_mean_slowhttptest = slowhttptest_df[' Fwd IAT Mean']\n",
    "print(\"Statistics for 'Fwd IAT Mean' under 'DoS Slowhttptest'\")\n",
    "print(f\"Mean: {fwd_iat_mean_slowhttptest.mean()}\")\n",
    "print(f\"Max: {fwd_iat_mean_slowhttptest.max()}\")\n",
    "print(f\"Std: {fwd_iat_mean_slowhttptest.std()}\")\n",
    "\n",
    "# For Non-'DoS Slowhttptest' (assuming dfs[0] is 'BENIGN' and others are various types of attacks)\n",
    "non_slowhttptest_dfs = [df for i, df in enumerate(dfs) if i != slowhttptest_index]\n",
    "non_slowhttptest_destination_port = [df[' Destination Port'] for df in non_slowhttptest_dfs]\n",
    "non_slowhttptest_fwd_iat_mean = [df[' Fwd IAT Mean'] for df in non_slowhttptest_dfs]\n",
    "\n",
    "# Stats for Non-'DoS Slowhttptest'\n",
    "print(\"Statistics for Non-'DoS Slowhttptest'\")\n",
    "print(\"For 'Destination Port'\")\n",
    "print(f\"Unique Ports: {[df.unique() for df in non_slowhttptest_destination_port]}\")\n",
    "print(\"For 'Fwd IAT Mean'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_slowhttptest_fwd_iat_mean]}\")\n",
    "print(f\"Max: {[df.max() for df in non_slowhttptest_fwd_iat_mean]}\")\n",
    "print(f\"Std: {[df.std() for df in non_slowhttptest_fwd_iat_mean]}\")\n",
    "\n",
    "# Visualization using matplotlib\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "# Histogram for 'Fwd IAT Mean' under 'DoS Slowhttptest'\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(fwd_iat_mean_slowhttptest, bins=50, color='blue', alpha=0.7, label='DoS Slowhttptest')\n",
    "plt.xlabel('Fwd IAT Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Histogram of Fwd IAT Mean for DoS Slowhttptest')\n",
    "plt.legend()\n",
    "\n",
    "# Histogram for 'Fwd IAT Mean' for Non-'DoS Slowhttptest'\n",
    "plt.subplot(1, 2, 2)\n",
    "for i, df in enumerate(non_slowhttptest_fwd_iat_mean):\n",
    "    plt.hist(df, bins=50, alpha=0.5, label=labels_per_group[i])\n",
    "plt.xlabel('Fwd IAT Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Histogram of Fwd IAT Mean for Non-DoS Slowhttptest')\n",
    "plt.legend(loc='upper right')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ac16781-d854-46c0-82a5-cfc6439c7099",
   "metadata": {},
   "source": [
    "The information provided shows statistical summaries for a dataset that presumably deals with network traffic, specifically focusing on \"Destination Port\", \"Fwd IAT Mean\", and the label \"DoS Slowhttptest\". \n",
    "\n",
    "#### Summary of the Output\n",
    "\n",
    "- For traffic labeled \"DoS Slowhttptest\":\n",
    "  - The 'Destination Port' is consistently 80, a standard port for HTTP traffic. \n",
    "  - The 'Fwd IAT Mean' (Forward Inter-Arrival Time Mean) has a mean value of around 12,803,522, a maximum value of 36,700,000, and a standard deviation of around 10,242,335. \n",
    "  \n",
    "- For traffic not labeled \"DoS Slowhttptest\":\n",
    "  - The 'Destination Port' has a wide range of values, which could include ports for various services (HTTP, DNS, NTP, etc.). \n",
    "\n",
    "#### Interpretations\n",
    "\n",
    "- **Destination Port 80 for \"DoS Slowhttptest\"**: Since this is a known attack that generally targets web servers, it is not surprising to see port 80 being the only destination port. This port is commonly used for HTTP traffic.\n",
    "  \n",
    "- **Fwd IAT Mean for \"DoS Slowhttptest\"**: This is a measure of the mean time between the forwarding of packets in the traffic. The high values could indicate slower or delayed packet forwarding, which is characteristic of slow HTTP DoS attacks. The standard deviation also shows high variability in the inter-arrival times, which could be indicative of an attack pattern.\n",
    "\n",
    "- **Wide Range of Ports for Non-\"DoS Slowhttptest\"**: The wide range of destination ports for the non-DoS traffic likely indicates a mix of different types of legitimate traffic.\n",
    "\n",
    "#### Considerations\n",
    "\n",
    "This basic analysis could be extended in various ways for more insight:\n",
    "\n",
    "- For \"Fwd IAT Mean\", further analysis like median or quartile information could provide more robust statistics.\n",
    "  \n",
    "- For \"Destination Port\" under non-DoS traffic, a frequency analysis could reveal the most commonly used ports, aiding in the differentiation of standard traffic from potentially malicious activity.\n",
    "  \n",
    "- Additional features and labels could be analyzed for a more comprehensive view of the data.\n",
    "\n",
    "This data alone cannot confirm a DoS attack but can provide valuable indicators that could be combined with other types of analyses and data for more conclusive results."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "027a2565-ef52-490a-89da-76052da8a922",
   "metadata": {},
   "source": [
    "Given that your dataset has 4,000 features, 100,000 samples, and a binary target variable, you would need to select machine learning models that are both effective at classification and capable of handling a large feature space without overfitting. Here are the models I recommend, in order of priority:\n",
    "\n",
    "### 1. Random Forest\n",
    "- **Argument**: Random Forest models are ensemble learning methods that are good for dealing with high-dimensionality. They can capture complex interactions between features without requiring feature scaling. Importantly, Random Forests have built-in feature selection and provide measures of feature importances.\n",
    "- **Evaluation**: Given the high dimensionality of your data, Random Forest can effectively perform dimensionality reduction. Additionally, they are less prone to overfitting, which is a concern when you have more features than samples. Their ability to parallelize can also make computation more manageable.\n",
    "\n",
    "### 2. Gradient Boosting Machines (GBM), e.g., XGBoost, LightGBM\n",
    "- **Argument**: GBMs, particularly implementations like XGBoost and LightGBM, are designed for speed and performance. These models perform well for a wide variety of classification tasks and are highly customizable.\n",
    "- **Evaluation**: They can handle large data sets efficiently but are also prone to overfitting if not properly tuned. You have a large enough dataset to mitigate this risk. These algorithms handle imbalanced classes quite well, and they also provide feature importance metrics.\n",
    "\n",
    "### 3. Support Vector Machines (SVM)\n",
    "- **Argument**: SVMs are effective in high-dimensional spaces and are also effective when the number of dimensions is greater than the number of samples. They are memory efficient and provide flexibility through the kernel trick.\n",
    "- **Evaluation**: The primary concern with SVM is the computational complexity, particularly when the dataset is large. However, they are less prone to overfitting especially when using radial basis function (RBF) or polynomial kernels.\n",
    "\n",
    "### 4. Logistic Regression with L1 or L2 Regularization\n",
    "- **Argument**: Logistic Regression is simple but effective for binary classification problems. Regularization techniques like L1 or L2 can be added to prevent overfitting.\n",
    "- **Evaluation**: Given the high number of features, using L1 regularization can help in feature selection by driving less important feature coefficients to zero. It's computationally cheaper but might not capture complex relationships between features as well as ensemble methods.\n",
    "\n",
    "### 5. Neural Networks\n",
    "- **Argument**: Deep learning has proven effective for complex pattern recognition but generally requires a large amount of data to be effective.\n",
    "- **Evaluation**: In a problem with a high number of features but a relatively moderate number of samples, neural networks might easily overfit. However, with proper architecture design and regularization techniques (such as dropout), they might prove effective.\n",
    "\n",
    "### 6. k-Nearest Neighbors (k-NN)\n",
    "- **Argument**: k-NN is a simple, non-parametric method that can be surprisingly effective for classification tasks.\n",
    "- **Evaluation**: The algorithm can be very slow for large datasets and high dimensions due to the curse of dimensionality. It's likely not the best choice for your dataset unless dimensionality can be significantly reduced beforehand.\n",
    "\n",
    "### 7. Naive Bayes\n",
    "- **Argument**: Naive Bayes classifiers are simple and fast, suitable for high-dimensional datasets.\n",
    "- **Evaluation**: They make a strong assumption about the independence of the features, which is often not the case in real-world applications. Given your high-dimensional data, this may lead to poor performance.\n",
    "\n",
    "### Conclusion\n",
    "Given your specific requirements, Random Forest and Gradient Boosting Machines are the most promising candidates, followed by SVM and Logistic Regression. Neural Networks could also be a good option if configured and regularized appropriately."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e2018c4-07d6-42e3-bc9b-5405def5d8ae",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'SSH-Patator':\n",
    "if ['Destination Port'] == 22 and ['Fwd Packet Length Mean'] > threshold:\n",
    "    return 'SSH-Patator'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "c226b33f-dee1-4183-a609-d4f7e3165b76",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'SSH-Patator' is at index 11 in labels_per_group\n",
      "Statistics for 'Destination Port' under 'SSH-Patator'\n",
      "Mean: 22.0\n",
      "Max: 22\n",
      "Std: 0.0\n",
      "Statistics for 'Fwd Packet Length Mean' under 'SSH-Patator'\n",
      "Mean: 48.10517343858373\n",
      "Max: 174.1818182\n",
      "Std: 47.792422381220995\n",
      "Statistics for Non-'SSH-Patator'\n",
      "For 'Destination Port'\n",
      "Mean: [9407.82391272463, 17560.41114701131, 81.94824935528665, 80.0, 80.0, 80.0, 80.0, 21.0, 444.0, 444.0, 8629.93484144819, 80.0, 80.0, 80.0]\n",
      "Max: [65534, 53938, 64873, 80, 80, 80, 80, 21, 444, 444, 65389, 80, 80, 80]\n",
      "Std: [19745.242209782715, 19017.78880711812, 336.9055571454257, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 13475.6892963097, 0.0, 0.0, 0.0]\n",
      "For 'Fwd Packet Length Mean'\n",
      "Mean: [66.43078926458519, 116.21841861694669, 7.40588951368109, 59.20787826049969, 44.579436380856166, 158.2734136135335, 63.51428676409833, 9.359669444457, 5.1522990821666665, 301.98209193181816, 1.0080582605077655, 17.219615082086406, 62.18333333333333, 8.535335342682925]\n",
      "Max: [4672.0, 5675.444444, 10.0, 398.0625, 317.25, 1983.0, 239.0, 15.0, 7.443968594, 920.75, 147.3, 216.5073892, 134.25, 241.3054187]\n",
      "Std: [204.2573453007653, 616.4774016077347, 1.2361810644729843, 56.251254207751614, 39.534729901827326, 407.3398286028672, 78.70837439201327, 2.4926717706871613, 1.1508994041850165, 187.87939733601297, 1.340298636782255, 53.72818435672608, 65.3384347884617, 43.07720889365892]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Find the index for 'SSH-Patator'\n",
    "ssh_patator_index = labels_per_group.index('SSH-Patator')\n",
    "print(f\"'SSH-Patator' is at index {ssh_patator_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'SSH-Patator' DataFrame\n",
    "ssh_patator_df = dfs[ssh_patator_index]\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Destination Port'\n",
    "dest_port_ssh_patator = ssh_patator_df[' Destination Port']\n",
    "print(\"Statistics for 'Destination Port' under 'SSH-Patator'\")\n",
    "print(f\"Mean: {dest_port_ssh_patator.mean()}\")\n",
    "print(f\"Max: {dest_port_ssh_patator.max()}\")\n",
    "print(f\"Std: {dest_port_ssh_patator.std()}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Fwd Packet Length Mean'\n",
    "fwd_packet_len_mean_ssh_patator = ssh_patator_df[' Fwd Packet Length Mean']\n",
    "print(\"Statistics for 'Fwd Packet Length Mean' under 'SSH-Patator'\")\n",
    "print(f\"Mean: {fwd_packet_len_mean_ssh_patator.mean()}\")\n",
    "print(f\"Max: {fwd_packet_len_mean_ssh_patator.max()}\")\n",
    "print(f\"Std: {fwd_packet_len_mean_ssh_patator.std()}\")\n",
    "\n",
    "# For Non-'SSH-Patator'\n",
    "non_ssh_patator_dfs = [df for i, df in enumerate(dfs) if i != ssh_patator_index]\n",
    "non_ssh_patator_dest_port = [df[' Destination Port'] for df in non_ssh_patator_dfs]\n",
    "non_ssh_patator_fwd_packet_len_mean = [df[' Fwd Packet Length Mean'] for df in non_ssh_patator_dfs]\n",
    "\n",
    "# Stats for Non-'SSH-Patator'\n",
    "print(\"Statistics for Non-'SSH-Patator'\")\n",
    "print(\"For 'Destination Port'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_ssh_patator_dest_port]}\")\n",
    "print(f\"Max: {[df.max() for df in non_ssh_patator_dest_port]}\")\n",
    "print(f\"Std: {[df.std() for df in non_ssh_patator_dest_port]}\")\n",
    "\n",
    "print(\"For 'Fwd Packet Length Mean'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_ssh_patator_fwd_packet_len_mean]}\")\n",
    "print(f\"Max: {[df.max() for df in non_ssh_patator_fwd_packet_len_mean]}\")\n",
    "print(f\"Std: {[df.std() for df in non_ssh_patator_fwd_packet_len_mean]}\")\n",
    "\n",
    "# Visualization\n",
    "plt.figure(figsize=(14, 6))\n",
    "\n",
    "# For 'Destination Port'\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(dest_port_ssh_patator, alpha=0.5, label='SSH-Patator', bins=20)\n",
    "plt.hist([df.mean() for df in non_ssh_patator_dest_port], alpha=0.5, label='Non-SSH-Patator', bins=20)\n",
    "plt.xlabel('Destination Port')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Distribution of Destination Port')\n",
    "plt.legend()\n",
    "\n",
    "# For 'Fwd Packet Length Mean'\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.hist(fwd_packet_len_mean_ssh_patator, alpha=0.5, label='SSH-Patator', bins=20)\n",
    "plt.hist([df.mean() for df in non_ssh_patator_fwd_packet_len_mean], alpha=0.5, label='Non-SSH-Patator', bins=20)\n",
    "plt.xlabel('Fwd Packet Length Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Distribution of Fwd Packet Length Mean')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b387b65c-2a99-4ec1-ac73-56dccaab4fd3",
   "metadata": {},
   "source": [
    "### Evaluation of the Heuristic\n",
    "The heuristic `if ['Destination Port'] == 22 and ['Fwd Packet Length Mean'] > threshold: return 'SSH-Patator'` seems promising based on the statistics. Specifically:\n",
    "\n",
    "1. `'Destination Port'` for `'SSH-Patator'` is **consistently 22**, while for Non-SSH-Patator cases, the mean Destination Port varies significantly and typically is not 22.\n",
    "2. `'Fwd Packet Length Mean'` for `'SSH-Patator'` has a mean value of **48.10** and a standard deviation of **47.79**. For Non-SSH-Patator cases, the mean and standard deviation vary considerably across different attack types. Hence, choosing a threshold for this feature could help to distinguish SSH-Patator effectively.\n",
    "\n",
    "Given that the two features appear to be good discriminators for the `'SSH-Patator'` label, the heuristic could be quite effective if the threshold is chosen wisely (considering mean and standard deviation).\n",
    "\n",
    "### Machine Learning Models\n",
    "1. **Decision Trees / Random Forest**: These models are good at capturing complex relationships between features and can work well for classification tasks like this one. Given that you've identified some key features (`'Destination Port'` and `'Fwd Packet Length Mean'`), a decision tree or a Random Forest (ensemble of decision trees) can easily take this into account. The model is also interpretable, which is a bonus.\n",
    "  \n",
    "2. **Logistic Regression**: This model could be effective, especially if the relationship between the labels and features is approximately linear after some transformation. Logistic Regression has the advantage of being simpler and faster to train. If the heuristic is close to an optimal solution, logistic regression might capture this relationship effectively.\n",
    "  \n",
    "3. **Support Vector Machines (SVM)**: SVMs are effective in high-dimensional spaces and are capable of creating complex decision boundaries. Given the high dimensionality of the dataset, SVM could be very effective.\n",
    "\n",
    "4. **K-Nearest Neighbors (K-NN)**: This model could work well if similar kinds of attacks cluster together in the feature space. However, K-NN may not be the best option here due to its sensitivity to dimensionality and the need for feature scaling.\n",
    "\n",
    "5. **Neural Networks**: Deep learning could potentially capture the complex relationships between different features, but it might be overkill for this problem and could risk overfitting unless regularized properly.\n",
    "\n",
    "6. **Naive Bayes**: Given that the features appear to have different distributions for different kinds of labels, a Naive Bayes classifier could be a good probabilistic model for this problem. However, the assumption of feature independence might not hold, limiting its effectiveness.\n",
    "\n",
    "### Prioritized List Based on Effectiveness\n",
    "1. Decision Trees / Random Forest\n",
    "2. Logistic Regression\n",
    "3. Support Vector Machines (SVM)\n",
    "4. Neural Networks\n",
    "5. K-Nearest Neighbors (K-NN)\n",
    "6. Naive Bayes\n",
    "\n",
    "The prioritization is based on the capacity of the models to handle feature importance effectively, their interpretability, and the complexity of the model in relation to the problem at hand."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e30b6ea9-6dbf-4411-895c-7f6a0a1f3d73",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'Web Attack – XSS':\n",
    "if ['Destination Port'] in [80, 443] and ['Fwd Packet Length Max'] > threshold:\n",
    "    return 'Web Attack – XSS'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "f6f6da14-36b2-4fc3-b712-964d396f3f8f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Web Attack – XSS' is at index 14 in labels_per_group\n",
      "Statistics for 'Destination Port' under 'Web Attack – XSS'\n",
      "Mean: 80.0\n",
      "Max: 80\n",
      "Std: 0.0\n",
      "Statistics for 'Fwd Packet Length Max' under 'Web Attack – XSS'\n",
      "Mean: 22.28048780487805\n",
      "Max: 585\n",
      "Std: 110.90248523831782\n",
      "Statistics for Non-'Web Attack – XSS'\n",
      "For 'Destination Port'\n",
      "Mean: [9407.82391272463, 17560.41114701131, 81.94824935528665, 80.0, 80.0, 80.0, 80.0, 21.0, 444.0, 444.0, 8629.93484144819, 22.0, 80.0, 80.0]\n",
      "Max: [65534, 53938, 64873, 80, 80, 80, 80, 21, 444, 444, 65389, 22, 80, 80]\n",
      "Std: [19745.242209782715, 19017.78880711812, 336.9055571454257, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 13475.6892963097, 0.0, 0.0, 0.0]\n",
      "For 'Fwd Packet Length Max'\n",
      "Mean: [230.6553349304018, 408.7205169628433, 14.932255504860146, 311.76727328809375, 233.66139223043814, 235.63481524249423, 94.67981374965763, 18.9382, 5309.333333333333, 1023.1363636363636, 1.0695330836454433, 323.50242326332796, 54.9072708113804, 277.6666666666667]\n",
      "Max: [24820, 23360, 20, 791, 423, 1983, 410, 49, 5792, 1460, 397, 1432, 602, 600]\n",
      "Std: [791.7018215043946, 2271.518192395181, 6.728071781624097, 199.62902808262837, 164.22856224473418, 427.33344976254864, 111.5918255736925, 5.572198007799761, 747.7439847077786, 409.2625534572369, 3.629565217016018, 321.0237363319064, 165.9429831921441, 290.95120907225333]\n"
     ]
    }
   ],
   "source": [
    "# Find the index for 'Web Attack – XSS'\n",
    "xss_index = labels_per_group.index('Web Attack � XSS')\n",
    "print(f\"'Web Attack – XSS' is at index {xss_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'Web Attack – XSS' DataFrame\n",
    "xss_df = dfs[xss_index]\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Destination Port'\n",
    "dest_port_xss = xss_df[' Destination Port']\n",
    "print(\"Statistics for 'Destination Port' under 'Web Attack – XSS'\")\n",
    "print(f\"Mean: {dest_port_xss.mean()}\")\n",
    "print(f\"Max: {dest_port_xss.max()}\")\n",
    "print(f\"Std: {dest_port_xss.std()}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for 'Fwd Packet Length Max'\n",
    "fwd_pkt_len_max_xss = xss_df[' Fwd Packet Length Max']\n",
    "print(\"Statistics for 'Fwd Packet Length Max' under 'Web Attack – XSS'\")\n",
    "print(f\"Mean: {fwd_pkt_len_max_xss.mean()}\")\n",
    "print(f\"Max: {fwd_pkt_len_max_xss.max()}\")\n",
    "print(f\"Std: {fwd_pkt_len_max_xss.std()}\")\n",
    "\n",
    "# For Non-'Web Attack – XSS' \n",
    "non_xss_dfs = [df for i, df in enumerate(dfs) if i != xss_index]\n",
    "non_xss_dest_port = [df[' Destination Port'] for df in non_xss_dfs]\n",
    "non_xss_fwd_pkt_len_max = [df[' Fwd Packet Length Max'] for df in non_xss_dfs]\n",
    "\n",
    "# Stats for Non-'Web Attack – XSS'\n",
    "print(\"Statistics for Non-'Web Attack – XSS'\")\n",
    "print(\"For 'Destination Port'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_xss_dest_port]}\")\n",
    "print(f\"Max: {[df.max() for df in non_xss_dest_port]}\")\n",
    "print(f\"Std: {[df.std() for df in non_xss_dest_port]}\")\n",
    "\n",
    "print(\"For 'Fwd Packet Length Max'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_xss_fwd_pkt_len_max]}\")\n",
    "print(f\"Max: {[df.max() for df in non_xss_fwd_pkt_len_max]}\")\n",
    "print(f\"Std: {[df.std() for df in non_xss_fwd_pkt_len_max]}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "ad1b8490-1131-47f7-8e28-eb5ac0c47b45",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "# Visualization for 'Destination Port' for Web Attack – XSS vs. Others\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.title(\"Histogram of 'Destination Port' for Web Attack – XSS\")\n",
    "plt.hist(xss_df[' Destination Port'], bins=20, alpha=0.5, label='Web Attack – XSS', color='b')\n",
    "plt.xlabel(\"Destination Port\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.title(\"Histogram of 'Destination Port' for Non-'Web Attack – XSS'\")\n",
    "plt.hist(np.concatenate(non_xss_dest_port), bins=20, alpha=0.5, label='Non-Web Attack – XSS', color='r')\n",
    "plt.xlabel(\"Destination Port\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Visualization for 'Fwd Packet Length Max' for Web Attack – XSS vs. Others\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.title(\"Histogram of 'Fwd Packet Length Max' for Web Attack – XSS\")\n",
    "plt.hist(xss_df[' Fwd Packet Length Max'], bins=20, alpha=0.5, label='Web Attack – XSS', color='b')\n",
    "plt.xlabel(\"'Fwd Packet Length Max'\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.title(\"Histogram of 'Fwd Packet Length Max' for Non-'Web Attack – XSS'\")\n",
    "plt.hist(np.concatenate(non_xss_fwd_pkt_len_max), bins=20, alpha=0.5, label='Non-Web Attack – XSS', color='r')\n",
    "plt.xlabel(\"'Fwd Packet Length Max'\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58843aa7-737e-43f8-b8cb-4d75b4ec425f",
   "metadata": {},
   "source": [
    "\n",
    "### Evaluation of the Heuristic\n",
    "\n",
    "Based on the statistics, we see the following:\n",
    "\n",
    "1. For 'Web Attack – XSS', the mean 'Destination Port' is 80, and the mean 'Fwd Packet Length Max' is 22.28 with a max value of 585.\n",
    "2. For Non-'Web Attack – XSS', the mean 'Destination Port' ranges from 21 to 17560, and the mean 'Fwd Packet Length Max' ranges from 1 to 5309.\n",
    "\n",
    "The heuristic \"if ['Destination Port'] in [80, 443] and ['Fwd Packet Length Max'] > threshold\" can be quite effective at identifying 'Web Attack – XSS' cases.\n",
    "\n",
    "### Machine Learning Models\n",
    "\n",
    "1. **Random Forest Classifier**: Random forests are often highly accurate and good at handling unbalanced datasets. Given the range and variability of features, this could be an effective model.\n",
    "\n",
    "2. **Gradient Boosting**: Like Random Forests, Gradient Boosting algorithms can capture complex patterns and are less likely to overfit.\n",
    "\n",
    "3. **Support Vector Machines (SVM)**: For a high-dimensional space, SVM could work well, especially if the heuristic condition creates clear boundaries.\n",
    "\n",
    "4. **Logistic Regression**: Despite its simplicity, logistic regression can be very powerful if the heuristic provides a strong linear boundary between the classes.\n",
    "\n",
    "5. **k-Nearest Neighbors (k-NN)**: Given that similar attack vectors might share similar features, k-NN could also be effective.\n",
    "\n",
    "### Prioritization Based on Data Statistics\n",
    "\n",
    "1. **Random Forest Classifier**: Most versatile, can handle the high dimensionality, and different ranges of feature values.\n",
    "  \n",
    "2. **Gradient Boosting**: Good for unbalanced classes, and can build complex decision boundaries.\n",
    "  \n",
    "3. **SVM**: Given the high-dimensional feature space, but computationally expensive.\n",
    "  \n",
    "4. **Logistic Regression**: Simpler but might work well if heuristic is strong.\n",
    "  \n",
    "5. **k-NN**: Could work but might suffer due to the high dimensionality and computational cost.\n",
    "\n",
    "The heuristic, if combined with one of these machine learning models, could improve the identification accuracy of 'Web Attack – XSS'."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81b74db9-2a1a-4175-b37c-5cd7ae4532da",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'Web Attack - Brute Force':\n",
    "if ['Destination Port'] in [80, 443] and ['Fwd Packet Length Mean'] > threshold:\n",
    "    return 'Web Attack - Brute Force'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "c90d82db-92b0-4ddf-8614-4a5ca4cf2033",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Web Attack – Brute Force' is at index 12 in labels_per_group\n",
      "Statistics for 'Destination Port' under 'Web Attack – Brute Force'\n",
      "Mean: 80.0\n",
      "Max: 80\n",
      "Std: 0.0\n",
      "Statistics for 'Fwd Packet Length Mean' under 'Web Attack – Brute Force'\n",
      "Mean: 17.219615082086406\n",
      "Max: 216.5073892\n",
      "Std: 53.72818435672608\n",
      "Statistics for Non-'Web Attack – Brute Force'\n",
      "For 'Destination Port'\n",
      "Mean: [9407.82391272463, 17560.41114701131, 81.94824935528665, 80.0, 80.0, 80.0, 80.0, 21.0, 444.0, 444.0, 8629.93484144819, 22.0, 80.0, 80.0]\n",
      "Max: [65534, 53938, 64873, 80, 80, 80, 80, 21, 444, 444, 65389, 22, 80, 80]\n",
      "Std: [19745.242209782715, 19017.78880711812, 336.9055571454257, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 13475.6892963097, 0.0, 0.0, 0.0]\n",
      "For 'Fwd Packet Length Mean'\n",
      "Mean: [66.43078926458519, 116.21841861694669, 7.40588951368109, 59.20787826049969, 44.579436380856166, 158.2734136135335, 63.51428676409833, 9.359669444457, 5.1522990821666665, 301.98209193181816, 1.0080582605077655, 48.10517343858373, 62.18333333333333, 8.535335342682925]\n",
      "Max: [4672.0, 5675.444444, 10.0, 398.0625, 317.25, 1983.0, 239.0, 15.0, 7.443968594, 920.75, 147.3, 174.1818182, 134.25, 241.3054187]\n",
      "Std: [204.2573453007653, 616.4774016077347, 1.2361810644729843, 56.251254207751614, 39.534729901827326, 407.3398286028672, 78.70837439201327, 2.4926717706871613, 1.1508994041850165, 187.87939733601297, 1.340298636782255, 47.792422381220995, 65.3384347884617, 43.07720889365892]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Find the index for 'Web Attack – Brute Force'\n",
    "bruteforce_index = labels_per_group.index('Web Attack � Brute Force')\n",
    "print(f\"'Web Attack – Brute Force' is at index {bruteforce_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'Web Attack – Brute Force' DataFrame\n",
    "bruteforce_df = dfs[bruteforce_index]\n",
    "\n",
    "# Calculate statistics for 'Destination Port'\n",
    "dest_port_bruteforce = bruteforce_df[' Destination Port']\n",
    "print(\"Statistics for 'Destination Port' under 'Web Attack – Brute Force'\")\n",
    "print(f\"Mean: {dest_port_bruteforce.mean()}\")\n",
    "print(f\"Max: {dest_port_bruteforce.max()}\")\n",
    "print(f\"Std: {dest_port_bruteforce.std()}\")\n",
    "\n",
    "# Calculate statistics for 'Fwd Packet Length Mean'\n",
    "fwd_pkt_len_mean_bruteforce = bruteforce_df[' Fwd Packet Length Mean']\n",
    "print(\"Statistics for 'Fwd Packet Length Mean' under 'Web Attack – Brute Force'\")\n",
    "print(f\"Mean: {fwd_pkt_len_mean_bruteforce.mean()}\")\n",
    "print(f\"Max: {fwd_pkt_len_mean_bruteforce.max()}\")\n",
    "print(f\"Std: {fwd_pkt_len_mean_bruteforce.std()}\")\n",
    "\n",
    "# Filter based on your conditions and calculate the statistics for non-'Web Attack – Brute Force'\n",
    "non_bruteforce_dfs = [df for i, df in enumerate(dfs) if i != bruteforce_index]\n",
    "non_bruteforce_dest_port = [df[' Destination Port'] for df in non_bruteforce_dfs]\n",
    "non_bruteforce_fwd_pkt_len_mean = [df[' Fwd Packet Length Mean'] for df in non_bruteforce_dfs]\n",
    "\n",
    "# Stats for Non-'Web Attack – Brute Force'\n",
    "print(\"Statistics for Non-'Web Attack – Brute Force'\")\n",
    "print(\"For 'Destination Port'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_bruteforce_dest_port]}\")\n",
    "print(f\"Max: {[df.max() for df in non_bruteforce_dest_port]}\")\n",
    "print(f\"Std: {[df.std() for df in non_bruteforce_dest_port]}\")\n",
    "\n",
    "print(\"For 'Fwd Packet Length Mean'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_bruteforce_fwd_pkt_len_mean]}\")\n",
    "print(f\"Max: {[df.max() for df in non_bruteforce_fwd_pkt_len_mean]}\")\n",
    "print(f\"Std: {[df.std() for df in non_bruteforce_fwd_pkt_len_mean]}\")\n",
    "\n",
    "# Visualization\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.title('Destination Port Distribution')\n",
    "plt.hist(dest_port_bruteforce, alpha=0.5, label='Web Attack – Brute Force', bins=30)\n",
    "plt.hist([df[' Destination Port'] for df in non_bruteforce_dfs], alpha=0.5, label='Non-Web Attack – Brute Force', bins=30)\n",
    "plt.xlabel('Destination Port')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.title('Fwd Packet Length Mean Distribution')\n",
    "plt.hist(fwd_pkt_len_mean_bruteforce, alpha=0.5, label='Web Attack – Brute Force', bins=30)\n",
    "plt.hist([df[' Fwd Packet Length Mean'] for df in non_bruteforce_dfs], alpha=0.5, label='Non-Web Attack – Brute Force', bins=30)\n",
    "plt.xlabel('Fwd Packet Length Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b41e3407-1d7a-4499-9798-9e6059e689e9",
   "metadata": {},
   "source": [
    "### Evaluating the Heuristic\n",
    "Given the output from the preceding code, it appears that the heuristic \"`if ['Destination Port'] in [80, 443] and ['Fwd Packet Length Mean'] > threshold: return 'Web Attack - Brute Force'`\" could be somewhat effective but limited in scope. The 'Destination Port' is 80, and 'Fwd Packet Length Mean' has a standard deviation of 53.72, indicating a high variance. The high standard deviation means that there could be many false positives if the threshold isn't set carefully. \n",
    "\n",
    "### Machine Learning Models for Distinguishing Cases\n",
    "Based on the statistics and the nature of the problem (classification), here are machine learning models that could be useful, prioritized by the technique most likely to be effective:\n",
    "\n",
    "1. **Random Forest Classifier**\n",
    "   - **Why**: Random Forests are usually very robust and can handle a variety of data distributions. They perform well on imbalanced datasets and can capture complex feature interactions. \n",
    "   - **Evaluation**: Given the high variance in 'Fwd Packet Length Mean' and the mixed nature of the 'Destination Port' (mostly 80 but could be others), a Random Forest Classifier can potentially create complex decision boundaries that would be difficult to model with simpler algorithms.\n",
    "\n",
    "2. **Gradient Boosting Machines (XGBoost, LightGBM)**\n",
    "   - **Why**: Boosting algorithms are known for high performance and can be fine-tuned for specific loss functions, which is particularly useful for imbalanced classification problems. \n",
    "   - **Evaluation**: The complexity in the data and the high standard deviation in one of the key features ('Fwd Packet Length Mean') makes gradient boosting a good candidate for capturing the underlying patterns effectively.\n",
    "\n",
    "3. **Support Vector Machines (SVM)**\n",
    "   - **Why**: SVMs are effective when the classes are not linearly separable, and they work well in high dimensional spaces.\n",
    "   - **Evaluation**: Given that the dataset might have a lot of features and the classes may not be linearly separable, SVM could be useful. However, SVMs might struggle with the large dataset and imbalanced classes.\n",
    "\n",
    "4. **Logistic Regression**\n",
    "   - **Why**: This is a simple yet effective algorithm for binary classification problems and serves as a good baseline.\n",
    "   - **Evaluation**: Logistic Regression may not capture the complexity in the feature interaction but can provide a solid baseline to compare with more complex models.\n",
    "\n",
    "5. **Neural Networks**\n",
    "   - **Why**: Deep Learning techniques can capture complex relations but may be an overkill for this problem unless the dataset is large enough.\n",
    "   - **Evaluation**: Depending on the size and complexity of the data, a simple neural network could be effective but would require much more computational resources and fine-tuning compared to other algorithms.\n",
    "\n",
    "### Summary\n",
    "While the heuristic could serve as a basic preliminary filter, machine learning models like Random Forest or Gradient Boosting Machines would likely provide a more accurate and robust method for distinguishing between 'Web Attack - Brute Force' and other cases. Given the high standard deviation in 'Fwd Packet Length Mean', a model that can capture complex patterns and handle varied distributions in the data would likely be the most effective."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d5965fb-b10c-4b36-8df7-58dfdf237ae7",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'Web Attack – SQL Injection':\n",
    "if ['Destination Port'] in [80, 443] and ['Fwd Packet Length Std'] > threshold:\n",
    "    return 'Web Attack – SQL Injection'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "4b1d2932-6094-492b-9c5d-486beb703a42",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Statistics for 'Fwd Packet Length Std' under 'Web Attack – Sql Injection'\n",
      "Mean: 131.19690015\n",
      "Max: 268.5\n",
      "Std: 137.23933331556995\n",
      "Statistics for Non-'Web Attack – Sql Injection'\n",
      "For 'Fwd Packet Length Std'\n",
      "Mean: 68.97449471089139\n",
      "Max: 7125.5968458437\n",
      "Std: 280.71749950473117\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# Check if label exists\n",
    "try:\n",
    "    sql_injection_index = labels_per_group.index('Web Attack � Sql Injection')\n",
    "except ValueError:\n",
    "    print(\"'Web Attack – Sql Injection' label not found in labels_per_group.\")\n",
    "    sql_injection_index = None\n",
    "\n",
    "if sql_injection_index is not None:\n",
    "    # Check if DataFrame exists at the index\n",
    "    if len(dfs) > sql_injection_index:\n",
    "        sql_injection_df = dfs[sql_injection_index]\n",
    "    else:\n",
    "        print(\"DataFrame not found for 'Web Attack – Sql Injection'\")\n",
    "        sql_injection_df = None\n",
    "\n",
    "    if sql_injection_df is not None:\n",
    "        # Check if column exists\n",
    "        if ' Fwd Packet Length Std' in sql_injection_df.columns:\n",
    "            fwd_packet_length_std_sql = sql_injection_df[' Fwd Packet Length Std']\n",
    "            \n",
    "            # Check if column is empty or filled with NaNs\n",
    "            if not fwd_packet_length_std_sql.empty and not fwd_packet_length_std_sql.isna().all():\n",
    "                print(\"Statistics for 'Fwd Packet Length Std' under 'Web Attack – Sql Injection'\")\n",
    "                print(f\"Mean: {fwd_packet_length_std_sql.mean()}\")\n",
    "                print(f\"Max: {fwd_packet_length_std_sql.max()}\")\n",
    "                print(f\"Std: {fwd_packet_length_std_sql.std()}\")\n",
    "            else:\n",
    "                print(\"Column ' Fwd Packet Length Std' is empty or filled with NaNs for 'Web Attack – Sql Injection'\")\n",
    "        else:\n",
    "            print(\"Column ' Fwd Packet Length Std' not found in DataFrame for 'Web Attack – Sql Injection'\")\n",
    "\n",
    "    # For Non-'Web Attack – Sql Injection'\n",
    "    non_sql_injection_dfs = [df for i, df in enumerate(dfs) if i != sql_injection_index]\n",
    "    non_sql_fwd_packet_length_std = [df[' Fwd Packet Length Std'] for df in non_sql_injection_dfs if ' Fwd Packet Length Std' in df.columns]\n",
    "\n",
    "    # Stats for Non-'Web Attack – Sql Injection', check if list is empty or filled with NaNs\n",
    "    if non_sql_fwd_packet_length_std:\n",
    "        non_sql_fwd_packet_length_std = pd.concat(non_sql_fwd_packet_length_std)\n",
    "        if not non_sql_fwd_packet_length_std.empty and not non_sql_fwd_packet_length_std.isna().all():\n",
    "            print(\"Statistics for Non-'Web Attack – Sql Injection'\")\n",
    "            print(\"For 'Fwd Packet Length Std'\")\n",
    "            print(f\"Mean: {non_sql_fwd_packet_length_std.mean()}\")\n",
    "            print(f\"Max: {non_sql_fwd_packet_length_std.max()}\")\n",
    "            print(f\"Std: {non_sql_fwd_packet_length_std.std()}\")\n",
    "        else:\n",
    "            print(\"Column ' Fwd Packet Length Std' is empty or filled with NaNs for Non-'Web Attack – Sql Injection'\")\n",
    "    else:\n",
    "        print(\"Column ' Fwd Packet Length Std' not found in DataFrames for Non-'Web Attack – Sql Injection'\")\n",
    "\n",
    "# Matplotlib Visualization, only if data is available\n",
    "if 'fwd_packet_length_std_sql' in locals() and 'non_sql_fwd_packet_length_std' in locals():\n",
    "    if not fwd_packet_length_std_sql.empty and not fwd_packet_length_std_sql.isna().all() and not non_sql_fwd_packet_length_std.empty and not non_sql_fwd_packet_length_std.isna().all():\n",
    "        plt.figure(figsize=(12, 6))\n",
    "\n",
    "        # Histogram for 'Fwd Packet Length Std' under 'Web Attack – Sql Injection'\n",
    "        plt.subplot(1, 2, 1)\n",
    "        plt.hist(fwd_packet_length_std_sql.dropna(), bins=20, color='blue', alpha=0.7, label='Web Attack – Sql Injection')\n",
    "        plt.title('Histogram of Fwd Packet Length Std for Web Attack – Sql Injection')\n",
    "        plt.xlabel('Fwd Packet Length Std')\n",
    "        plt.ylabel('Frequency')\n",
    "\n",
    "        # Histogram for 'Fwd Packet Length Std' for Non-'Web Attack – Sql Injection'\n",
    "        plt.subplot(1, 2, 2)\n",
    "        plt.hist(non_sql_fwd_packet_length_std.dropna(), bins=20, color='red', alpha=0.7, label='Non-Web Attack – Sql Injection')\n",
    "        plt.title('Histogram of Fwd Packet Length Std for Non-Web Attack – Sql Injection')\n",
    "        plt.xlabel('Fwd Packet Length Std')\n",
    "        plt.ylabel('Frequency')\n",
    "\n",
    "        plt.tight_layout()\n",
    "        plt.show()\n",
    "    else:\n",
    "        print(\"Histograms cannot be generated due to lack of data or all NaN values.\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8ee6130-78a1-4f65-8bcc-fef6b38634d5",
   "metadata": {},
   "source": [
    "### Heuristic Evaluation\n",
    "\n",
    "Given the data statistics, the heuristic \"if ['Destination Port'] in [80, 443] and ['Fwd Packet Length Std'] > threshold: return 'Web Attack – SQL Injection'\" seems to make some intuitive sense. The mean \"Fwd Packet Length Std\" for \"Web Attack – SQL Injection\" (131.20) is higher than for other labels (68.97), suggesting that there's a difference between the two cases that could be captured by this feature. However, the standard deviation for non-'Web Attack – SQL Injection' cases is quite high (280.72), which indicates that there might be significant overlap, potentially leading to false positives or negatives.\n",
    "\n",
    "### Machine Learning Models\n",
    "\n",
    "Based on the problem and data, here are some machine learning models that could be used to distinguish between 'Web Attack – SQL Injection' and other labels:\n",
    "\n",
    "1. **Random Forest Classifier**\n",
    "    - **Argument**: Handles a mix of numerical and categorical data well. Built-in feature importance can provide insight into which features are crucial for classification. Random Forests are robust to outliers, which is beneficial given the high standard deviation in the non-SQL Injection cases.\n",
    "    - **Evaluation**: Random Forests are good at capturing complex patterns and could possibly learn the high variance in non-SQL Injection cases, making it the top choice.\n",
    "\n",
    "2. **Gradient Boosting Classifier**\n",
    "    - **Argument**: Like Random Forests, Gradient Boosting works well for both categorical and numerical features. It's excellent for imbalanced datasets and often provides high accuracy.\n",
    "    - **Evaluation**: The algorithm could adjust to the different mean and std deviations between SQL Injection and non-SQL Injection classes effectively but might be sensitive to outliers.\n",
    "    \n",
    "3. **Support Vector Machine (SVM)**\n",
    "    - **Argument**: SVM works well for a clear margin of separation and is effective in high dimensional spaces.\n",
    "    - **Evaluation**: Given that 'Fwd Packet Length Std' already shows some degree of separation (based on mean), SVM might be effective. However, the high standard deviation in non-SQL Injection cases might be a challenge.\n",
    "\n",
    "4. **Logistic Regression**\n",
    "    - **Argument**: Simple and fast. If the heuristic works reasonably well, the relationship between the label and features may not be overly complex, making logistic regression a suitable choice.\n",
    "    - **Evaluation**: It's less likely to capture complex relationships compared to ensemble methods, but it might work well if the problem is indeed linearly separable.\n",
    "\n",
    "5. **k-Nearest Neighbors (k-NN)**\n",
    "    - **Argument**: If the cases are clustered in a multi-dimensional space, k-NN could be effective.\n",
    "    - **Evaluation**: The high standard deviation in the non-SQL Injection cases could make this less effective because the 'neighborhood' might be less distinguishable.\n",
    "\n",
    "6. **Neural Networks**\n",
    "    - **Argument**: Can capture complex, non-linear relationships.\n",
    "    - **Evaluation**: Might be overkill for this problem and require a lot of data. Also, interpretability could be a challenge.\n",
    "\n",
    "### Prioritization\n",
    "I would start with the Random Forest Classifier given its robustness and ability to handle both categorical and numerical features effectively, followed by Gradient Boosting for its capability to adapt well to imbalanced data."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c338c8ae-6306-460a-99ee-963e1f4c08ae",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'Bot':\n",
    "if ['Flow IAT Mean'] < threshold and ['Fwd Packets/s'] > threshold:\n",
    "    return 'Bot'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "f3fa56a8-a064-4c88-976e-5fc5d19bfc26",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Bot' is at index 1 in labels_per_group\n",
      "Statistics for 'Flow IAT Mean' under 'Bot'\n",
      "Mean: 61107.206150487815\n",
      "Max: 3541331.765\n",
      "Std: 131362.42364759522\n",
      "Statistics for 'Fwd Packets/s' under 'Bot'\n",
      "Mean: 21909.755234135733\n",
      "Max: 1000000.0\n",
      "Std: 72616.15479380582\n",
      "Statistics for Non-'Bot'\n",
      "For 'Flow IAT Mean'\n",
      "Mean: [907627.6790029698, 1888593.3395309653, 14142488.398697596, 4801625.845836498, 9616745.909722473, 10343177.448566975, 194217.76507314, 24513.994875, 1390563.4338931818, 24968.677645501855, 120526.33184754077, 1550640.3584925712, 351473.8988083333, 1673592.2127546342]\n",
      "Max: [120000000.0, 40700000.0, 119000000.0, 58700000.0, 28300000.0, 57900000.0, 468701.0435, 24843.10537, 3613460.517, 39600000.0, 4130073.448, 1999045.667, 719802.5714, 1997993.333]\n",
      "Std: [4118823.6973127592, 2570284.474247773, 26663308.595294688, 4036434.087524249, 5147457.206511686, 13457172.961114958, 196652.3534525864, 226.02254799776125, 1297978.1561337472, 733126.9546316357, 148503.43530999921, 643519.5594609973, 367699.4685450803, 454855.84656894964]\n",
      "For 'Fwd Packets/s'\n",
      "Mean: [58336.44929953641, 110.52567466844931, 8.402375795992906, 180434.16712970092, 11520.914323777937, 2267.239656366967, 114975.01260700004, 23.463037056666668, 4552.569026617091, 31338.01166634469, 7824.46186999524, 1713.5656554089821, 8038.566174310584, 1834.2213117334634]\n",
      "Max: [3000000.0, 1500000.0, 2587.322122, 3000000.0, 1000000.0, 500000.0, 2000000.0, 23.51222893, 100000.0, 1000000.0, 1000000.0, 50000.0, 23809.52381, 500000.0]\n",
      "Std: [231401.64974840987, 7316.509489861263, 108.61162085405643, 442206.809140501, 71789.52527484816, 11759.058063592704, 303831.14302315755, 0.0408124061593136, 21318.485434522125, 63781.258494291775, 20868.570963942635, 6157.231458959035, 9323.198053848027, 25193.607828152708]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Find the index for 'Bot'\n",
    "bot_index = labels_per_group.index('Bot')\n",
    "print(f\"'Bot' is at index {bot_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'Bot' DataFrame\n",
    "bot_df = dfs[bot_index]\n",
    "\n",
    "# Calculate the statistics for 'Flow IAT Mean'\n",
    "flow_iat_mean_bot = bot_df[' Flow IAT Mean']\n",
    "print(\"Statistics for 'Flow IAT Mean' under 'Bot'\")\n",
    "print(f\"Mean: {flow_iat_mean_bot.mean()}\")\n",
    "print(f\"Max: {flow_iat_mean_bot.max()}\")\n",
    "print(f\"Std: {flow_iat_mean_bot.std()}\")\n",
    "\n",
    "# Calculate the statistics for 'Fwd Packets/s'\n",
    "fwd_packets_per_s_bot = bot_df['Fwd Packets/s']\n",
    "print(\"Statistics for 'Fwd Packets/s' under 'Bot'\")\n",
    "print(f\"Mean: {fwd_packets_per_s_bot.mean()}\")\n",
    "print(f\"Max: {fwd_packets_per_s_bot.max()}\")\n",
    "print(f\"Std: {fwd_packets_per_s_bot.std()}\")\n",
    "\n",
    "# For Non-'Bot'\n",
    "non_bot_dfs = [df for i, df in enumerate(dfs) if i != bot_index]\n",
    "non_bot_flow_iat_mean = [df[' Flow IAT Mean'] for df in non_bot_dfs]\n",
    "non_bot_fwd_packets_per_s = [df['Fwd Packets/s'] for df in non_bot_dfs]\n",
    "\n",
    "# Stats for Non-'Bot'\n",
    "print(\"Statistics for Non-'Bot'\")\n",
    "print(\"For 'Flow IAT Mean'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_bot_flow_iat_mean]}\")\n",
    "print(f\"Max: {[df.max() for df in non_bot_flow_iat_mean]}\")\n",
    "print(f\"Std: {[df.std() for df in non_bot_flow_iat_mean]}\")\n",
    "\n",
    "print(\"For 'Fwd Packets/s'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_bot_fwd_packets_per_s]}\")\n",
    "print(f\"Max: {[df.max() for df in non_bot_fwd_packets_per_s]}\")\n",
    "print(f\"Std: {[df.std() for df in non_bot_fwd_packets_per_s]}\")\n",
    "\n",
    "# Visualization using Matplotlib\n",
    "# Bot\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(flow_iat_mean_bot, bins=20, alpha=0.5, label='Bot - Flow IAT Mean', color='r')\n",
    "plt.title('Histogram of Flow IAT Mean for Bot')\n",
    "plt.xlabel('Flow IAT Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.grid(True)\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.hist(fwd_packets_per_s_bot, bins=20, alpha=0.5, label='Bot - Fwd Packets/s', color='b')\n",
    "plt.title('Histogram of Fwd Packets/s for Bot')\n",
    "plt.xlabel('Fwd Packets/s')\n",
    "plt.ylabel('Frequency')\n",
    "plt.grid(True)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Non-Bot\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist([item for sublist in non_bot_flow_iat_mean for item in sublist], bins=20, alpha=0.5, label='Non-Bot - Flow IAT Mean', color='r')\n",
    "plt.title('Histogram of Flow IAT Mean for Non-Bot')\n",
    "plt.xlabel('Flow IAT Mean')\n",
    "plt.ylabel('Frequency')\n",
    "plt.grid(True)\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.hist([item for sublist in non_bot_fwd_packets_per_s for item in sublist], bins=20, alpha=0.5, label='Non-Bot - Fwd Packets/s', color='b')\n",
    "plt.title('Histogram of Fwd Packets/s for Non-Bot')\n",
    "plt.xlabel('Fwd Packets/s')\n",
    "plt.ylabel('Frequency')\n",
    "plt.grid(True)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cdb5cc28-a054-43e8-8c28-ee990ef73a03",
   "metadata": {},
   "source": [
    "### Heuristic Evaluation:\n",
    "\n",
    "To apply the heuristic `if ['Flow IAT Mean'] < threshold and ['Fwd Packets/s'] > threshold: return 'Bot'`, we'll need to choose appropriate threshold values for `Flow IAT Mean` and `Fwd Packets/s`. Based on the statistics:\n",
    "\n",
    "- For `Bot` label, the mean of `Flow IAT Mean` is approximately 61,107, and the mean of `Fwd Packets/s` is approximately 21,910.\n",
    "- For `Non-Bot`, the mean of `Flow IAT Mean` ranges between 24,514 and 14,142,488, and the mean of `Fwd Packets/s` ranges between 8.40 and 180,434.\n",
    "\n",
    "The heuristic could be effective if we choose a threshold for `Flow IAT Mean` that is greater than 61,107 but less than the minimum mean of `Non-Bot` labels (i.e., 24,514). Likewise, for `Fwd Packets/s`, a threshold greater than 8.40 but less than 21,910 might work.\n",
    "\n",
    "### Machine Learning Models:\n",
    "\n",
    "#### 1. Random Forest Classifier\n",
    "\n",
    "- **Rationale**: Random Forests are great for handling imbalanced classes and can model complex relationships between features.\n",
    "- **Evaluation**: Given the wide variance in the statistics (both within `Bot` and between `Bot` and `Non-Bot`), a Random Forest model may be able to accurately capture these variances and generalize well.\n",
    "\n",
    "#### 2. Gradient Boosting Machines (XGBoost, LightGBM)\n",
    "\n",
    "- **Rationale**: Like Random Forests, these algorithms are also ensemble methods but are known for higher performance and speed.\n",
    "- **Evaluation**: These models can capture the complex patterns in the data but may be prone to overfitting if not carefully tuned. Considering our data has high variance and potentially many outliers (based on the `Max` and `Std` values), gradient boosting could work well.\n",
    "\n",
    "#### 3. Support Vector Machines (SVM)\n",
    "\n",
    "- **Rationale**: SVMs are effective in high-dimensional spaces and can model non-linear relationships via the kernel trick.\n",
    "- **Evaluation**: The `Std` values are high, which means the data points are spread out over a wide range. If the data isn't linearly separable in the original feature space, the kernel trick could be useful.\n",
    "\n",
    "#### 4. Logistic Regression\n",
    "\n",
    "- **Rationale**: This is a simple but effective algorithm for binary classification problems.\n",
    "- **Evaluation**: Given the high variance and potentially non-linear relationships between features, logistic regression may not be the best model but can serve as a good baseline model.\n",
    "\n",
    "#### 5. Neural Networks\n",
    "\n",
    "- **Rationale**: Capable of modeling highly complex relationships between features.\n",
    "- **Evaluation**: They require a large amount of data and are computationally expensive. The wide variance and large `Max` values in the statistics might make the network susceptible to outliers and overfitting.\n",
    "\n",
    "#### Prioritization:\n",
    "\n",
    "1. Random Forest Classifier\n",
    "2. Gradient Boosting Machines\n",
    "3. Support Vector Machines\n",
    "4. Logistic Regression\n",
    "5. Neural Networks\n",
    "\n",
    "The Random Forest Classifier is prioritized the highest because it can handle the complexities and imbalances in the dataset while being relatively easier to tune. Gradient Boosting Machines come next for their performance and ability to model complex patterns. SVM is useful for high dimensions but might require more computational effort. Logistic Regression is a good baseline but might be too simple, and Neural Networks come last due to their susceptibility to overfitting and computational costs.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55d5e606-9d33-4d8b-9cca-0996340f5415",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'Infiltration':\n",
    "if ['Total Length of Fwd Packets'] > threshold and ['Total Length of Bwd Packets'] > threshold:\n",
    "    return 'Infiltration'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "185f2f41-e123-4c38-b16c-c32b5e795422",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Infiltration' is at index 9 in labels_per_group\n",
      "Statistics for 'Total Length of Fwd Packets' under 'Infiltration'\n",
      "Mean: 333898.5909090909\n",
      "Max: 1827335\n",
      "Std: 615938.5818042373\n",
      "Statistics for 'Total Length of Bwd Packets' under 'Infiltration'\n",
      "Mean: 4776.681818181818\n",
      "Max: 22866\n",
      "Std: 7976.809885007819\n",
      "Statistics for Non-'Infiltration'\n",
      "For 'Total Length of Fwd Packets'\n",
      "Mean: [628.4507155378931, 2717.592084006462, 31.93803312834755, 415.54040715607647, 282.87495171614614, 489.34699769053117, 818.8863325116406, 59.4936, 14420.333333333334, 1.0935830212234707, 1014.3058696822833, 2289.245521601686, 277.6666666666667, 1421.121951219512]\n",
      "Max: [1288022, 205731, 62, 6910, 2538, 5720, 3146, 135, 20858, 1473, 3832, 43951, 600, 48985]\n",
      "Std: [5899.544998591592, 18225.779771668746, 11.976004716877853, 515.662249159252, 267.62670184403936, 1092.9414957674119, 1149.311616938412, 46.30769309227038, 3232.5936129780785, 6.375544548212053, 1007.6277485498973, 9698.299219058885, 290.95120907225333, 8102.912360594667]\n",
      "For 'Total Length of Bwd Packets'\n",
      "Mean: [18726.075606757975, 63.10823909531502, 7402.54477038286, 6506.773442319556, 7798.953164661737, 111.2924364896074, 18.420706655710763, 92.8512, 7867718.5, 12.291146067415731, 1389.5245018847604, 3751.1559536354057, 1025.1666666666667, 5344.578048780488]\n",
      "Max: [639650620, 626, 11613, 11680, 13043, 3705, 9375, 188, 7879536, 11595, 5673, 72214, 4149, 183697]\n",
      "Std: [2475577.676500454, 81.41096825732993, 5576.459372614145, 5210.758776366451, 5441.152487165844, 445.91985762353784, 191.86909335485893, 93.90593976775234, 27113.41202246593, 268.08988624339867, 1383.7162748924322, 15943.20073584664, 1329.0414750579616, 30448.41548753756]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Find the index for 'Infiltration'\n",
    "infiltration_index = labels_per_group.index('Infiltration')\n",
    "print(f\"'Infiltration' is at index {infiltration_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'Infiltration' DataFrame\n",
    "infiltration_df = dfs[infiltration_index]\n",
    "\n",
    "# Calculate the statistics for 'Total Length of Fwd Packets'\n",
    "fwd_packets_length_infiltration = infiltration_df['Total Length of Fwd Packets']\n",
    "print(\"Statistics for 'Total Length of Fwd Packets' under 'Infiltration'\")\n",
    "print(f\"Mean: {fwd_packets_length_infiltration.mean()}\")\n",
    "print(f\"Max: {fwd_packets_length_infiltration.max()}\")\n",
    "print(f\"Std: {fwd_packets_length_infiltration.std()}\")\n",
    "\n",
    "# Calculate the statistics for 'Total Length of Bwd Packets'\n",
    "bwd_packets_length_infiltration = infiltration_df[' Total Length of Bwd Packets']\n",
    "print(\"Statistics for 'Total Length of Bwd Packets' under 'Infiltration'\")\n",
    "print(f\"Mean: {bwd_packets_length_infiltration.mean()}\")\n",
    "print(f\"Max: {bwd_packets_length_infiltration.max()}\")\n",
    "print(f\"Std: {bwd_packets_length_infiltration.std()}\")\n",
    "\n",
    "# For Non-'Infiltration'\n",
    "non_infiltration_dfs = [df for i, df in enumerate(dfs) if i != infiltration_index]\n",
    "non_infiltration_fwd_packets_length = [df['Total Length of Fwd Packets'] for df in non_infiltration_dfs]\n",
    "non_infiltration_bwd_packets_length = [df[' Total Length of Bwd Packets'] for df in non_infiltration_dfs]\n",
    "\n",
    "# Stats for Non-'Infiltration'\n",
    "print(\"Statistics for Non-'Infiltration'\")\n",
    "print(\"For 'Total Length of Fwd Packets'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_infiltration_fwd_packets_length]}\")\n",
    "print(f\"Max: {[df.max() for df in non_infiltration_fwd_packets_length]}\")\n",
    "print(f\"Std: {[df.std() for df in non_infiltration_fwd_packets_length]}\")\n",
    "\n",
    "print(\"For 'Total Length of Bwd Packets'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_infiltration_bwd_packets_length]}\")\n",
    "print(f\"Max: {[df.max() for df in non_infiltration_bwd_packets_length]}\")\n",
    "print(f\"Std: {[df.std() for df in non_infiltration_bwd_packets_length]}\")\n",
    "\n",
    "# Visualize the data using matplotlib\n",
    "plt.figure(figsize=(12, 6))\n",
    "\n",
    "# Plot for 'Total Length of Fwd Packets'\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.hist(fwd_packets_length_infiltration, bins=30, alpha=0.5, label='Infiltration')\n",
    "for df in non_infiltration_fwd_packets_length:\n",
    "    plt.hist(df, bins=30, alpha=0.3, label='Non-Infiltration')\n",
    "plt.xlabel('Total Length of Fwd Packets')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Total Length of Fwd Packets Distribution')\n",
    "plt.legend()\n",
    "\n",
    "# Plot for 'Total Length of Bwd Packets'\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.hist(bwd_packets_length_infiltration, bins=30, alpha=0.5, label='Infiltration')\n",
    "for df in non_infiltration_bwd_packets_length:\n",
    "    plt.hist(df, bins=30, alpha=0.3, label='Non-Infiltration')\n",
    "plt.xlabel('Total Length of Bwd Packets')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Total Length of Bwd Packets Distribution')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ced7082a-2601-4280-8fdb-6eda86521550",
   "metadata": {},
   "source": [
    "### Evaluating the Heuristic\n",
    "\n",
    "Based on the statistics for the 'Total Length of Fwd Packets' and 'Total Length of Bwd Packets':\n",
    "\n",
    "#### For 'Infiltration' attacks:\n",
    "- Mean of 'Total Length of Fwd Packets': 333898.6\n",
    "- Max of 'Total Length of Fwd Packets': 1827335\n",
    "- Std of 'Total Length of Fwd Packets': 615938.6\n",
    "\n",
    "- Mean of 'Total Length of Bwd Packets': 4776.7\n",
    "- Max of 'Total Length of Bwd Packets': 22866\n",
    "- Std of 'Total Length of Bwd Packets': 7976.8\n",
    "\n",
    "#### For Non-'Infiltration' activities:\n",
    "- The means for both 'Total Length of Fwd Packets' and 'Total Length of Bwd Packets' are much lower compared to those for 'Infiltration'.\n",
    "  \n",
    "A simple heuristic might be to set a threshold based on these statistics. For example, you could set the threshold for 'Total Length of Fwd Packets' to be greater than 600000 and for 'Total Length of Bwd Packets' to be greater than 8000 to identify 'Infiltration'.\n",
    "\n",
    "However, it's not a guarantee that this heuristic will work for all cases. Therefore, machine learning models might offer a more accurate and reliable solution.\n",
    "\n",
    "### Machine Learning Models\n",
    "\n",
    "1. **Random Forest Classifier**\n",
    "    - **Argument**: Random Forest can capture the complex relationships between the features and labels, and it is known for high accuracy and ability to handle imbalanced datasets.\n",
    "    - **Evaluation**: Given the variation in the data (as indicated by Std), Random Forest's ensemble approach would likely capture these nuances well.\n",
    "\n",
    "2. **Gradient Boosting Machines (e.g., XGBoost)**\n",
    "    - **Argument**: Similar to Random Forest, but generally performs even better albeit at the cost of increased computational intensity.\n",
    "    - **Evaluation**: Gradient Boosting would be adept at capturing the skewness in the data and providing high accuracy.\n",
    "\n",
    "3. **Support Vector Machines (SVM)**\n",
    "    - **Argument**: Effective in high-dimensional spaces and when classes are separable by a hyperplane.\n",
    "    - **Evaluation**: Given the statistics, it's unclear how well the two classes would be separated in the feature space, but it's worth trying.\n",
    "\n",
    "4. **Logistic Regression**\n",
    "    - **Argument**: If the relationship between the feature and the label is somewhat linear, logistic regression could work quite well and it's easy to interpret.\n",
    "    - **Evaluation**: Given the high Std and max values, the data may not be linear, but it is quick to implement and test.\n",
    "\n",
    "5. **Neural Networks**\n",
    "    - **Argument**: Can model complex non-linear relationships.\n",
    "    - **Evaluation**: Neural Networks may be overkill for this kind of problem, especially given the risk of overfitting and the computational resources needed.\n",
    "\n",
    "Based on your statistics, I would prioritize Random Forest and Gradient Boosting Machines because they can handle a wide range of data distributions and are less likely to overfit compared to a deep neural network."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72169bc3-3c5e-443d-9bb9-1dcea202989f",
   "metadata": {},
   "source": [
    "## See how well the following rule works\n",
    "\n",
    "'Heartbleed':\n",
    "if ['Destination Port'] == 443 and ['Fwd Packet Length Max'] > threshold:\n",
    "    return 'Heartbleed'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "b00a5525-8d1b-40ee-a807-a1e4b7e4ab49",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Heartbleed' is at index 8 in labels_per_group\n",
      "Statistics for 'Fwd Packet Length Max' under 'Heartbleed'\n",
      "Mean: 5309.333333333333\n",
      "Max: 5792\n",
      "Std: 747.7439847077786\n",
      "Statistics for Non-'Heartbleed'\n",
      "For 'Fwd Packet Length Max'\n",
      "Mean: [230.6553349304018, 408.7205169628433, 14.932255504860146, 311.76727328809375, 233.66139223043814, 235.63481524249423, 94.67981374965763, 18.9382, 1023.1363636363636, 1.0695330836454433, 323.50242326332796, 54.9072708113804, 277.6666666666667, 22.28048780487805]\n",
      "Max: [24820, 23360, 20, 791, 423, 1983, 410, 49, 1460, 397, 1432, 602, 600, 585]\n",
      "Std: [791.7018215043946, 2271.518192395181, 6.728071781624097, 199.62902808262837, 164.22856224473418, 427.33344976254864, 111.5918255736925, 5.572198007799761, 409.2625534572369, 3.629565217016018, 321.0237363319064, 165.9429831921441, 290.95120907225333, 110.90248523831782]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Find the index for 'Heartbleed'\n",
    "heartbleed_index = labels_per_group.index('Heartbleed')\n",
    "print(f\"'Heartbleed' is at index {heartbleed_index} in labels_per_group\")\n",
    "\n",
    "# Extract the 'Heartbleed' DataFrame\n",
    "heartbleed_df = dfs[heartbleed_index]\n",
    "\n",
    "# Filter and calculate statistics for 'Destination Port' and 'Fwd Packet Length Max'\n",
    "destination_port_heartbleed = heartbleed_df[' Destination Port']\n",
    "fwd_pkt_len_max_heartbleed = heartbleed_df[' Fwd Packet Length Max']\n",
    "\n",
    "# We will only look at statistics for 'Fwd Packet Length Max' as 'Destination Port' is constant in this case\n",
    "print(\"Statistics for 'Fwd Packet Length Max' under 'Heartbleed'\")\n",
    "print(f\"Mean: {fwd_pkt_len_max_heartbleed.mean()}\")\n",
    "print(f\"Max: {fwd_pkt_len_max_heartbleed.max()}\")\n",
    "print(f\"Std: {fwd_pkt_len_max_heartbleed.std()}\")\n",
    "\n",
    "# For Non-'Heartbleed'\n",
    "non_heartbleed_dfs = [df for i, df in enumerate(dfs) if i != heartbleed_index]\n",
    "non_heartbleed_fwd_pkt_len_max = [df[' Fwd Packet Length Max'] for df in non_heartbleed_dfs]\n",
    "\n",
    "# Stats for Non-'Heartbleed'\n",
    "print(\"Statistics for Non-'Heartbleed'\")\n",
    "print(\"For 'Fwd Packet Length Max'\")\n",
    "print(f\"Mean: {[df.mean() for df in non_heartbleed_fwd_pkt_len_max]}\")\n",
    "print(f\"Max: {[df.max() for df in non_heartbleed_fwd_pkt_len_max]}\")\n",
    "print(f\"Std: {[df.std() for df in non_heartbleed_fwd_pkt_len_max]}\")\n",
    "\n",
    "# Import matplotlib for visualization\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Histogram for 'Fwd Packet Length Max' under 'Heartbleed'\n",
    "plt.figure(figsize=(10, 6))\n",
    "plt.hist(fwd_pkt_len_max_heartbleed, bins=30, alpha=0.5, color='r', label='Heartbleed')\n",
    "plt.xlabel('Fwd Packet Length Max')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Distribution of Fwd Packet Length Max under Heartbleed')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# Histogram for 'Fwd Packet Length Max' under Non-'Heartbleed'\n",
    "plt.figure(figsize=(10, 6))\n",
    "for i, df in enumerate(non_heartbleed_fwd_pkt_len_max):\n",
    "    label = labels_per_group[i] if i < heartbleed_index else labels_per_group[i + 1]  # Skip the index corresponding to 'Heartbleed'\n",
    "    plt.hist(df, bins=30, alpha=0.3, label=label)\n",
    "plt.xlabel('Fwd Packet Length Max')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Distribution of Fwd Packet Length Max under Non-Heartbleed')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4257797c-97c8-4656-8f47-3a3283b4ca5f",
   "metadata": {},
   "source": [
    "### Evaluating the Heuristic:\n",
    "\n",
    "Given the output:\n",
    "\n",
    "- The mean value of 'Fwd Packet Length Max' for 'Heartbleed' attacks is about 5309.33, with a standard deviation of about 747.74.\n",
    "- The mean values for 'Fwd Packet Length Max' for Non-'Heartbleed' vary significantly, ranging from as low as 1.06 to as high as 1023.14.\n",
    "\n",
    "One can tentatively say that a threshold value somewhere between 1023 and 5309 may be an effective delimiter. However, without more information on the data distribution and overlapping values, it is hard to set a precise threshold confidently.\n",
    "\n",
    "### Machine Learning Models:\n",
    "\n",
    "1. **Random Forest Classifier**: \n",
    "    - **Argument**: This ensemble method is highly interpretable and can handle a mix of numerical and categorical features. It can also capture non-linear relationships.\n",
    "    - **Evaluation**: Given the large spread in 'Fwd Packet Length Max' across various labels (std ranges from 3.6 to 2271.5), a Random Forest could work well in capturing this complex distribution.\n",
    "\n",
    "2. **Gradient Boosting Machines (XGBoost, LightGBM)**:\n",
    "    - **Argument**: These algorithms are known for high performance and can optimize on a given evaluation metric, which could be crucial for imbalanced classes.\n",
    "    - **Evaluation**: Similar to Random Forests, but generally performs even better though at the cost of interpretability.\n",
    "\n",
    "3. **Support Vector Machines (SVM)**:\n",
    "    - **Argument**: Effective in high-dimensional spaces and best suited for problems with complex decision boundaries.\n",
    "    - **Evaluation**: Could be computationally expensive but may perform well given the range and standard deviation in the features.\n",
    "\n",
    "4. **Logistic Regression**:\n",
    "    - **Argument**: Provides good interpretability and works well if the relationship between the independent and dependent variables is approximately linear.\n",
    "    - **Evaluation**: Might be too simple to capture all the nuances in the data but could be used as a baseline model.\n",
    "\n",
    "5. **Neural Networks**:\n",
    "    - **Argument**: Can model highly complex relationships.\n",
    "    - **Evaluation**: May require a large amount of data and could overfit or be hard to interpret.\n",
    "\n",
    "6. **K-Nearest Neighbors (K-NN)**:\n",
    "    - **Argument**: Makes decisions based on the entire dataset and can capture non-linear decision boundaries.\n",
    "    - **Evaluation**: Given that the 'Fwd Packet Length Max' feature values have a wide range, normalization and distance metrics would be crucial for K-NN. \n",
    "\n",
    "### Prioritization:\n",
    "\n",
    "1. **Random Forest Classifier**: Given its balance of performance and interpretability, it's often a good starting point.\n",
    "2. **Gradient Boosting Machines**: For performance optimization.\n",
    "3. **SVM**: If computational complexity is not an issue.\n",
    "4. **Logistic Regression**: For a simpler, baseline model.\n",
    "5. **Neural Networks**: If the above methods are insufficient and more data is available.\n",
    "6. **K-NN**: Least preferred due to its sensitivity to feature scales and the potential for high computational cost.\n",
    "\n",
    "The best choice would depend on the specific needs of the project, including computational resources, the importance of interpretability, and performance requirements."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bc75563a-0e74-4e09-b37c-33cc7c9cf498",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}