Intrusion_Detection_on_UNSBW_NB15_by_using_Machine_Learning_Techniques.ipynb

{
  "metadata": {
    "kernelspec": {
      "language": "python",
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "version": "3.7.12",
      "mimetype": "text/x-python",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "file_extension": ".py"
    },
    "colab": {
      "name": "Intrusion Detection on UNSBW-NB15 by using Machine Learning Techniques.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "6uXtwVzPDwjb",
        "qUrBo1mqbYSv"
      ],
      "include_colab_link": true
    }
  },
  "nbformat_minor": 0,
  "nbformat": 4,
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/loaiwalid07/main/blob/main/Intrusion_Detection_on_UNSBW_NB15_by_using_Machine_Learning_Techniques.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <h1 style=' border:1;'><center>Intrusion Detection on UNSBW-NB15 by using Machine Learning Techniques</center></h1>\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "----"
      ],
      "metadata": {
        "id": "258cb14e"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Import Important Pacages"
      ],
      "metadata": {
        "id": "6uXtwVzPDwjb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install -q sklearn_som \n",
        "!pip install -q git+https://github.com/giuliano-oliveira/gdown_folder.git"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AzsVAwXvDwXE",
        "outputId": "9b58a1d5-4c35-4d4e-8810-157341f12a2c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "  Installing build dependencies ... \u001b[?25l\u001b[?25hdone\n",
            "  Getting requirements to build wheel ... \u001b[?25l\u001b[?25hdone\n",
            "    Preparing wheel metadata ... \u001b[?25l\u001b[?25hdone\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import random\n",
        "import gdown\n",
        "import os\n",
        "from collections import Counter\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import MinMaxScaler,StandardScaler,LabelEncoder,normalize,OneHotEncoder\n",
        "import plotly.graph_objects as go\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "from sklearn import svm\n",
        "from sklearn.ensemble import StackingClassifier,RandomForestClassifier,BaggingClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.inspection import permutation_importance\n",
        "from sklearn import preprocessing\n",
        "from sklearn.model_selection import cross_val_score,cross_validate, train_test_split, GridSearchCV, StratifiedKFold\n",
        "from sklearn import metrics\n",
        "from sklearn.metrics import plot_confusion_matrix, classification_report,roc_curve, roc_auc_score\n",
        "from sklearn.manifold import TSNE\n",
        "from sklearn_som.som import SOM\n",
        "from sklearn.pipeline import make_pipeline\n",
        "from xgboost import XGBClassifier\n",
        "from tqdm import trange"
      ],
      "metadata": {
        "id": "YfQCih75te38",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:16:22.849574Z",
          "iopub.execute_input": "2021-12-11T01:16:22.850520Z",
          "iopub.status.idle": "2021-12-11T01:16:52.029019Z",
          "shell.execute_reply.started": "2021-12-11T01:16:22.850480Z",
          "shell.execute_reply": "2021-12-11T01:16:52.027809Z"
        },
        "trusted": true
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Load the dataset**"
      ],
      "metadata": {
        "id": "oleywlFYY5aP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "for i in trange(1):\n",
        "  !gdown --id 1yafPDtNnSxCcYqhE5VhCMTUPk8PGfTXm\n",
        "  !gdown --id 19EBDKZ1WrFDSqEiIpmgUL3h84jqC2CbQ"
      ],
      "metadata": {
        "id": "MbJc2K6buSv_",
        "outputId": "8edb882c-8145-43b1-de3d-5c66cfcb72ac",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:04:55.517759Z",
          "iopub.execute_input": "2021-12-11T01:04:55.518061Z",
          "iopub.status.idle": "2021-12-11T01:04:59.279652Z",
          "shell.execute_reply.started": "2021-12-11T01:04:55.518031Z",
          "shell.execute_reply": "2021-12-11T01:04:59.278638Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\r  0%|          | 0/1 [00:00<?, ?it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1yafPDtNnSxCcYqhE5VhCMTUPk8PGfTXm\n",
            "To: /content/UNSW_NB15_.csv\n",
            "100% 15.4M/15.4M [00:00<00:00, 263MB/s]\n",
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=19EBDKZ1WrFDSqEiIpmgUL3h84jqC2CbQ\n",
            "To: /content/train-set_UNSW_NB15_.csv\n",
            "100% 32.3M/32.3M [00:00<00:00, 199MB/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 1/1 [00:04<00:00,  4.48s/it]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_set = pd.read_csv(\"./train-set_UNSW_NB15_.csv\") #, header = None\n",
        "test_set = pd.read_csv(\"./UNSW_NB15_.csv\") "
      ],
      "metadata": {
        "id": "fJuVGK1YCvts"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Data Cleaning & Prepocessing**"
      ],
      "metadata": {
        "id": "7uD1nQl4DW4t"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wQMuZMDNY7x0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "810014e7-3594-43ac-8a06-2ffb50f35b76"
      },
      "source": [
        "'''\n",
        "Checkng for nulls and drop duplicated values \n",
        "'''\n",
        "(train_set =='-').sum()\n",
        "\n",
        "test_set.isnull().sum().sum()\n",
        "\n",
        "print(Counter(train_set.attack_cat))\n",
        "\n",
        "bin_labels = pd.value_counts(train_set['label'], sort = True)\n",
        "bin_labels.rename(index={0:'Normal',1:'Attack'})\n",
        "multi_labels = pd.value_counts(train_set['attack_cat'], sort = True)\n",
        "\n",
        "train_set.duplicated().sum()\n",
        "\n",
        "train_set['service'].replace('-',np.nan,inplace=True)\n",
        "train_set.dropna(inplace=True)\n",
        "\n",
        "test_set['service'].replace('-',np.nan,inplace=True)\n",
        "test_set.dropna(inplace=True)\n",
        "\n",
        "train_set.isnull().sum().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Counter({'Normal': 56000, 'Generic': 40000, 'Exploits': 33393, 'Fuzzers': 18184, 'DoS': 12264, 'Reconnaissance': 10491, 'Analysis': 2000, 'Backdoor': 1746, 'Shellcode': 1133, 'Worms': 130})\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {},
          "execution_count": 93
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "duplicate = test_set[test_set.duplicated()]\n",
        "duplicate\n",
        "print(\"Duplicate Rows :\")\n",
        "  \n",
        "# Print the resultant Dataframe\n",
        "duplicate"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 87
        },
        "id": "oZrqCNIeEBly",
        "outputId": "cfadf245-96ff-4179-fbda-0c45dc99abcc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Duplicate Rows :\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>id</th>\n",
              "      <th>dur</th>\n",
              "      <th>proto</th>\n",
              "      <th>service</th>\n",
              "      <th>state</th>\n",
              "      <th>spkts</th>\n",
              "      <th>dpkts</th>\n",
              "      <th>sbytes</th>\n",
              "      <th>dbytes</th>\n",
              "      <th>rate</th>\n",
              "      <th>sttl</th>\n",
              "      <th>dttl</th>\n",
              "      <th>sload</th>\n",
              "      <th>dload</th>\n",
              "      <th>sloss</th>\n",
              "      <th>dloss</th>\n",
              "      <th>sinpkt</th>\n",
              "      <th>dinpkt</th>\n",
              "      <th>sjit</th>\n",
              "      <th>djit</th>\n",
              "      <th>swin</th>\n",
              "      <th>stcpb</th>\n",
              "      <th>dtcpb</th>\n",
              "      <th>dwin</th>\n",
              "      <th>tcprtt</th>\n",
              "      <th>synack</th>\n",
              "      <th>ackdat</th>\n",
              "      <th>smean</th>\n",
              "      <th>dmean</th>\n",
              "      <th>trans_depth</th>\n",
              "      <th>response_body_len</th>\n",
              "      <th>ct_srv_src</th>\n",
              "      <th>ct_state_ttl</th>\n",
              "      <th>ct_dst_ltm</th>\n",
              "      <th>ct_src_dport_ltm</th>\n",
              "      <th>ct_dst_sport_ltm</th>\n",
              "      <th>ct_dst_src_ltm</th>\n",
              "      <th>is_ftp_login</th>\n",
              "      <th>ct_ftp_cmd</th>\n",
              "      <th>ct_flw_http_mthd</th>\n",
              "      <th>ct_src_ltm</th>\n",
              "      <th>ct_srv_dst</th>\n",
              "      <th>is_sm_ips_ports</th>\n",
              "      <th>attack_cat</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "Empty DataFrame\n",
              "Columns: [id, dur, proto, service, state, spkts, dpkts, sbytes, dbytes, rate, sttl, dttl, sload, dload, sloss, dloss, sinpkt, dinpkt, sjit, djit, swin, stcpb, dtcpb, dwin, tcprtt, synack, ackdat, smean, dmean, trans_depth, response_body_len, ct_srv_src, ct_state_ttl, ct_dst_ltm, ct_src_dport_ltm, ct_dst_sport_ltm, ct_dst_src_ltm, is_ftp_login, ct_ftp_cmd, ct_flw_http_mthd, ct_src_ltm, ct_srv_dst, is_sm_ips_ports, attack_cat, label]\n",
              "Index: []"
            ]
          },
          "metadata": {},
          "execution_count": 94
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_set = train_set.drop([ 'stcpb', 'dtcpb', 'is_sm_ips_ports', 'ct_state_ttl', 'ct_flw_http_mthd', 'is_ftp_login', 'ct_ftp_cmd','id'], axis=1) # droped label\n",
        "test_set = test_set.drop([ 'stcpb', 'dtcpb', 'is_sm_ips_ports', 'ct_state_ttl', 'ct_flw_http_mthd', 'is_ftp_login', 'ct_ftp_cmd','id'], axis=1) # droped label\n",
        "\n",
        "\n",
        "# train_set.info()"
      ],
      "metadata": {
        "id": "5oLBEITYEv0B"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Data Insights**"
      ],
      "metadata": {
        "id": "Br_HG53OFDEY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Mutli Plot\n",
        "\n",
        "labels = ['NonAttacks','Attack']\n",
        "my_explode = [0.1,0,0,0,0,0,0,0,0,0]\n",
        "plt.subplot(1, 2, 1)\n",
        "multi_labels.plot(kind = 'pie', rot=0,figsize=(11, 6), explode=my_explode, autopct='%1.1f%%', startangle=15,)\n",
        "plt.title(\"Multi class distribution\")\n",
        "plt.xticks(range(2), labels)\n",
        "plt.xlabel(\"Class\")\n",
        "plt.ylabel(\"Frequency\")\n",
        "print('/n')\n",
        "\n",
        "# Binary Plot\n",
        "plt.subplot(1, 2, 2)\n",
        "bin_labels.plot(kind = 'pie', rot=0,figsize=(11, 6),autopct='%1.1f%%', startangle=15,)\n",
        "plt.title(\"Binary class distribution\")\n",
        "plt.xticks(range(2), labels)\n",
        "plt.xlabel(\"Class\")\n",
        "plt.ylabel(\"Frequency\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 396
        },
        "id": "HKQtn4zDE-YX",
        "outputId": "92ee63d8-69ec-4806-9c4f-57912fc2a178"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "/n\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Frequency')"
            ]
          },
          "metadata": {},
          "execution_count": 96
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 792x432 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_set.attack_cat.value_counts().plot.barh()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        },
        "id": "nZnq6E67GUoF",
        "outputId": "57dc4ca3-c078-4f2c-9e76-159c69a7b0d0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7ff94090ad90>"
            ]
          },
          "metadata": {},
          "execution_count": 97
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_set.attack_cat.value_counts()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jDmc04YvGkQl",
        "outputId": "71813d04-4698-488c-9d2c-eee09952de30"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Generic           39496\n",
              "Normal            19488\n",
              "Exploits          16187\n",
              "DoS                1791\n",
              "Fuzzers            1731\n",
              "Reconnaissance     1703\n",
              "Analysis            564\n",
              "Worms               114\n",
              "Backdoor             99\n",
              "Name: attack_cat, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 98
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import seaborn as sns\n",
        "corr = train_set.corr()\n",
        "\n",
        "sns.set_theme('notebook')\n",
        "\n",
        "mask = np.triu(np.ones_like(corr, dtype=bool))\n",
        "cmap = sns.diverging_palette(230, 20, as_cmap=True)\n",
        "\n",
        "plt.figure(figsize=(12,8))\n",
        "sns.heatmap(corr, mask=mask, cmap=cmap, center=0, square=True, linewidths=1.)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "id": "a6lt3xKzGZvf",
        "outputId": "3b555942-e214-4304-ecf7-a6fd02bd4bea"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7ff940be01d0>"
            ]
          },
          "metadata": {},
          "execution_count": 73
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x576 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        ""
      ],
      "metadata": {
        "id": "2cZrUS44T7vd"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#drop Normal Cat from train set\n",
        "remove = train_set[train_set['attack_cat']=='Normal'].index\n",
        "train_set = train_set.drop(remove)"
      ],
      "metadata": {
        "id": "EDifJngEGh_N"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#drop Normal Cat from test set\n",
        "\n",
        "remove = test_set[test_set['attack_cat']=='Normal'].index\n",
        "test_set = test_set.drop(remove)\n",
        "\n",
        "Y_test = test_set.attack_cat\n",
        "\n",
        "X_test = test_set.drop(['label'], axis=1) #, 'attack_cat'\n",
        "\n",
        "# transfer with label encoding\n",
        "X_test = X_test.apply(LabelEncoder().fit_transform)"
      ],
      "metadata": {
        "id": "ac1UgTNgG5Km"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        ""
      ],
      "metadata": {
        "id": "tSrp6EjjMtzr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pd.set_option('display.max_columns', None)\n",
        "\n",
        "isna = pd.DataFrame(train_set.isna().sum().values.reshape(1, 37), columns=train_set.columns, index=['NA'])\n",
        "isnull = pd.DataFrame(train_set.isnull().sum().values.reshape(1, 37), columns=train_set.columns, index=['Null'])\n",
        "\n",
        "print(pd.concat([isna, isnull]).sum().sum())\n",
        "pd.concat([isna, isnull])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 150
        },
        "id": "DZyaNgMeHIa9",
        "outputId": "5171c29a-6f59-4909-bf0f-140b8fbec844"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0\n"
          ]
        },
        {
          "output_type": "execute_result",
          "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>dur</th>\n",
              "      <th>proto</th>\n",
              "      <th>service</th>\n",
              "      <th>state</th>\n",
              "      <th>spkts</th>\n",
              "      <th>dpkts</th>\n",
              "      <th>sbytes</th>\n",
              "      <th>dbytes</th>\n",
              "      <th>rate</th>\n",
              "      <th>sttl</th>\n",
              "      <th>dttl</th>\n",
              "      <th>sload</th>\n",
              "      <th>dload</th>\n",
              "      <th>sloss</th>\n",
              "      <th>dloss</th>\n",
              "      <th>sinpkt</th>\n",
              "      <th>dinpkt</th>\n",
              "      <th>sjit</th>\n",
              "      <th>djit</th>\n",
              "      <th>swin</th>\n",
              "      <th>dwin</th>\n",
              "      <th>tcprtt</th>\n",
              "      <th>synack</th>\n",
              "      <th>ackdat</th>\n",
              "      <th>smean</th>\n",
              "      <th>dmean</th>\n",
              "      <th>trans_depth</th>\n",
              "      <th>response_body_len</th>\n",
              "      <th>ct_srv_src</th>\n",
              "      <th>ct_dst_ltm</th>\n",
              "      <th>ct_src_dport_ltm</th>\n",
              "      <th>ct_dst_sport_ltm</th>\n",
              "      <th>ct_dst_src_ltm</th>\n",
              "      <th>ct_src_ltm</th>\n",
              "      <th>ct_srv_dst</th>\n",
              "      <th>attack_cat</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>NA</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Null</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      dur  proto  service  state  spkts  dpkts  sbytes  dbytes  rate  sttl  \\\n",
              "NA      0      0        0      0      0      0       0       0     0     0   \n",
              "Null    0      0        0      0      0      0       0       0     0     0   \n",
              "\n",
              "      dttl  sload  dload  sloss  dloss  sinpkt  dinpkt  sjit  djit  swin  \\\n",
              "NA       0      0      0      0      0       0       0     0     0     0   \n",
              "Null     0      0      0      0      0       0       0     0     0     0   \n",
              "\n",
              "      dwin  tcprtt  synack  ackdat  smean  dmean  trans_depth  \\\n",
              "NA       0       0       0       0      0      0            0   \n",
              "Null     0       0       0       0      0      0            0   \n",
              "\n",
              "      response_body_len  ct_srv_src  ct_dst_ltm  ct_src_dport_ltm  \\\n",
              "NA                    0           0           0                 0   \n",
              "Null                  0           0           0                 0   \n",
              "\n",
              "      ct_dst_sport_ltm  ct_dst_src_ltm  ct_src_ltm  ct_srv_dst  attack_cat  \\\n",
              "NA                   0               0           0           0           0   \n",
              "Null                 0               0           0           0           0   \n",
              "\n",
              "      label  \n",
              "NA        0  \n",
              "Null      0  "
            ]
          },
          "metadata": {},
          "execution_count": 99
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Feature Selection**"
      ],
      "metadata": {
        "id": "qUrBo1mqbYSv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Preprocessing (transformation/scaling)\n",
        "\n",
        "le = LabelEncoder()\n",
        "le1 = LabelEncoder()\n",
        "le2 = LabelEncoder()\n",
        "le3 = LabelEncoder()\n",
        "\n",
        "vector = data['attack_cat']\n",
        "\n",
        "print(\"attack cat:\", set(list(vector))) # use print to make it print on single line \n",
        "\n",
        "data['attack_cat'] = le1.fit_transform(vector)\n",
        "data['proto'] =   le.fit_transform(data['proto'])\n",
        "data['service'] = le1.fit_transform(data['service'])\n",
        "data['state'] =   le2.fit_transform(data['state'])\n",
        "#data['srcip'] = le.fit_transform(data['srcip'])\n",
        "#data['dstip'] = le.fit_transform(data['dstip'])\n",
        "\n",
        "print('\\nDescribing attack_type: ')\n",
        "data['attack_cat'].describe()"
      ],
      "metadata": {
        "id": "DaQ1f4assN2T",
        "outputId": "ff103e82-4a5e-45e7-d2a3-b9a1b1e5f172",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:05:41.747053Z",
          "iopub.execute_input": "2021-12-11T01:05:41.747414Z",
          "iopub.status.idle": "2021-12-11T01:05:42.277789Z",
          "shell.execute_reply.started": "2021-12-11T01:05:41.747375Z",
          "shell.execute_reply": "2021-12-11T01:05:42.276756Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "attack cat: {'Shellcode', 'Generic', 'Analysis', 'Worms', 'Normal', 'Reconnaissance', 'DoS', 'Fuzzers', 'Exploits', 'Backdoor'}\n",
            "\n",
            "Describing attack_type: \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    82332.000000\n",
              "mean         4.987757\n",
              "std          1.431541\n",
              "min          0.000000\n",
              "25%          4.000000\n",
              "50%          5.000000\n",
              "75%          6.000000\n",
              "max          9.000000\n",
              "Name: attack_cat, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#To show the headers of the data\n",
        "data.columns"
      ],
      "metadata": {
        "id": "C2qwHtbRsVr6",
        "outputId": "1f1111a3-8e4f-46bc-b42a-88af89ddecc5",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:05:45.477256Z",
          "iopub.execute_input": "2021-12-11T01:05:45.477985Z",
          "iopub.status.idle": "2021-12-11T01:05:45.485500Z",
          "shell.execute_reply.started": "2021-12-11T01:05:45.477942Z",
          "shell.execute_reply": "2021-12-11T01:05:45.484612Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['id', 'dur', 'proto', 'service', 'state', 'spkts', 'dpkts', 'sbytes',\n",
              "       'dbytes', 'rate', 'sttl', 'dttl', 'sload', 'dload', 'sloss', 'dloss',\n",
              "       'sinpkt', 'dinpkt', 'sjit', 'djit', 'swin', 'stcpb', 'dtcpb', 'dwin',\n",
              "       'tcprtt', 'synack', 'ackdat', 'smean', 'dmean', 'trans_depth',\n",
              "       'response_body_len', 'ct_srv_src', 'ct_state_ttl', 'ct_dst_ltm',\n",
              "       'ct_src_dport_ltm', 'ct_dst_sport_ltm', 'ct_dst_src_ltm',\n",
              "       'is_ftp_login', 'ct_ftp_cmd', 'ct_flw_http_mthd', 'ct_src_ltm',\n",
              "       'ct_srv_dst', 'is_sm_ips_ports', 'attack_cat', 'label'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#X = data.drop(['id','attack_cat','label', 'stcpb', 'dtcpb', 'is_sm_ips_ports', 'ct_state_ttl', 'ct_flw_http_mthd', 'is_ftp_login', 'ct_ftp_cmd'], axis=1) # droped label\n",
        "#X = data[varr]\n",
        "\n",
        "X = data.drop(['id','attack_cat','label'], axis=1) \n",
        "y = data['attack_cat']\n",
        "X"
      ],
      "metadata": {
        "id": "JsINQEQzseL3",
        "outputId": "33cf4c85-63c6-44eb-bd98-1ee8e23d5e35",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:05:48.048183Z",
          "iopub.execute_input": "2021-12-11T01:05:48.048504Z",
          "iopub.status.idle": "2021-12-11T01:05:48.098146Z",
          "shell.execute_reply.started": "2021-12-11T01:05:48.048467Z",
          "shell.execute_reply": "2021-12-11T01:05:48.097211Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 443
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>dur</th>\n",
              "      <th>proto</th>\n",
              "      <th>service</th>\n",
              "      <th>state</th>\n",
              "      <th>spkts</th>\n",
              "      <th>dpkts</th>\n",
              "      <th>sbytes</th>\n",
              "      <th>dbytes</th>\n",
              "      <th>rate</th>\n",
              "      <th>sttl</th>\n",
              "      <th>dttl</th>\n",
              "      <th>sload</th>\n",
              "      <th>dload</th>\n",
              "      <th>sloss</th>\n",
              "      <th>dloss</th>\n",
              "      <th>sinpkt</th>\n",
              "      <th>dinpkt</th>\n",
              "      <th>sjit</th>\n",
              "      <th>djit</th>\n",
              "      <th>swin</th>\n",
              "      <th>stcpb</th>\n",
              "      <th>dtcpb</th>\n",
              "      <th>dwin</th>\n",
              "      <th>tcprtt</th>\n",
              "      <th>synack</th>\n",
              "      <th>ackdat</th>\n",
              "      <th>smean</th>\n",
              "      <th>dmean</th>\n",
              "      <th>trans_depth</th>\n",
              "      <th>response_body_len</th>\n",
              "      <th>ct_srv_src</th>\n",
              "      <th>ct_state_ttl</th>\n",
              "      <th>ct_dst_ltm</th>\n",
              "      <th>ct_src_dport_ltm</th>\n",
              "      <th>ct_dst_sport_ltm</th>\n",
              "      <th>ct_dst_src_ltm</th>\n",
              "      <th>is_ftp_login</th>\n",
              "      <th>ct_ftp_cmd</th>\n",
              "      <th>ct_flw_http_mthd</th>\n",
              "      <th>ct_src_ltm</th>\n",
              "      <th>ct_srv_dst</th>\n",
              "      <th>is_sm_ips_ports</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.000011</td>\n",
              "      <td>117</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>496</td>\n",
              "      <td>0</td>\n",
              "      <td>90909.090200</td>\n",
              "      <td>254</td>\n",
              "      <td>0</td>\n",
              "      <td>1.803636e+08</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.011000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>248</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.000008</td>\n",
              "      <td>117</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1762</td>\n",
              "      <td>0</td>\n",
              "      <td>125000.000300</td>\n",
              "      <td>254</td>\n",
              "      <td>0</td>\n",
              "      <td>8.810000e+08</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.008000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>881</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.000005</td>\n",
              "      <td>117</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1068</td>\n",
              "      <td>0</td>\n",
              "      <td>200000.005100</td>\n",
              "      <td>254</td>\n",
              "      <td>0</td>\n",
              "      <td>8.544000e+08</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.005000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>534</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.000006</td>\n",
              "      <td>117</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>900</td>\n",
              "      <td>0</td>\n",
              "      <td>166666.660800</td>\n",
              "      <td>254</td>\n",
              "      <td>0</td>\n",
              "      <td>6.000000e+08</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.006000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>450</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.000010</td>\n",
              "      <td>117</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>2126</td>\n",
              "      <td>0</td>\n",
              "      <td>100000.002500</td>\n",
              "      <td>254</td>\n",
              "      <td>0</td>\n",
              "      <td>8.504000e+08</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.010000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>1063</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>0</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",
              "      <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>82327</th>\n",
              "      <td>0.000005</td>\n",
              "      <td>117</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>104</td>\n",
              "      <td>0</td>\n",
              "      <td>200000.005100</td>\n",
              "      <td>254</td>\n",
              "      <td>0</td>\n",
              "      <td>8.320000e+07</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.005000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>52</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>82328</th>\n",
              "      <td>1.106101</td>\n",
              "      <td>111</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>20</td>\n",
              "      <td>8</td>\n",
              "      <td>18062</td>\n",
              "      <td>354</td>\n",
              "      <td>24.410067</td>\n",
              "      <td>254</td>\n",
              "      <td>252</td>\n",
              "      <td>1.241044e+05</td>\n",
              "      <td>2242.109863</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>55.880051</td>\n",
              "      <td>143.7</td>\n",
              "      <td>4798.130981</td>\n",
              "      <td>190.980813</td>\n",
              "      <td>255</td>\n",
              "      <td>1072535109</td>\n",
              "      <td>3284291478</td>\n",
              "      <td>255</td>\n",
              "      <td>0.173208</td>\n",
              "      <td>0.100191</td>\n",
              "      <td>0.073017</td>\n",
              "      <td>903</td>\n",
              "      <td>44</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>82329</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>6</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>46</td>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>60000.720000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>46</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>82330</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>6</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>46</td>\n",
              "      <td>0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.000000e+00</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>60000.732000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>10.954518</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>46</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>82331</th>\n",
              "      <td>0.000009</td>\n",
              "      <td>117</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>104</td>\n",
              "      <td>0</td>\n",
              "      <td>111111.107200</td>\n",
              "      <td>254</td>\n",
              "      <td>0</td>\n",
              "      <td>4.622222e+07</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.009000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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>0.000000</td>\n",
              "      <td>52</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>82332 rows × 42 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "            dur  proto  service  ...  ct_src_ltm  ct_srv_dst  is_sm_ips_ports\n",
              "0      0.000011    117        0  ...           1           2                0\n",
              "1      0.000008    117        0  ...           1           2                0\n",
              "2      0.000005    117        0  ...           1           3                0\n",
              "3      0.000006    117        0  ...           2           3                0\n",
              "4      0.000010    117        0  ...           2           3                0\n",
              "...         ...    ...      ...  ...         ...         ...              ...\n",
              "82327  0.000005    117        0  ...           2           1                0\n",
              "82328  1.106101    111        0  ...           3           2                0\n",
              "82329  0.000000      6        0  ...           1           1                1\n",
              "82330  0.000000      6        0  ...           1           1                1\n",
              "82331  0.000009    117        0  ...           1           1                0\n",
              "\n",
              "[82332 rows x 42 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_attck = []\n",
        "y_attck = []\n",
        "for i in range(len(y)):\n",
        "  if y[i] == 1 :\n",
        "    y_attck.append(y[i])\n",
        "    x_attck.append(X.iloc[i])"
      ],
      "metadata": {
        "id": "Lv1ribZ8gX-6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Binary Model**"
      ],
      "metadata": {
        "id": "00SscU6Xa-rA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Split the data set to use olny normal data\n",
        "\n",
        "num_col = train_set.select_dtypes(include='number').columns\n",
        "len(num_col)\n",
        "\n",
        "cat_col = train_set.columns.difference(num_col)\n",
        "cat_col = cat_col.drop(['attack_cat'])\n",
        "\n",
        "print(num_col.shape)\n",
        "train_cat = train_set[cat_col].copy()\n",
        "\n",
        "train_cat = pd.get_dummies(train_cat,columns=cat_col)\n",
        "print(train_cat.shape)\n",
        "\n",
        "\n",
        "train_num = train_set[num_col].copy()\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "data = pd.concat([train_set, train_cat],axis=1)\n",
        "print(data.shape)\n",
        "\n",
        "data.drop(columns=cat_col,inplace=True)\n",
        "print(data.shape)\n",
        "\n",
        "\n",
        "# data.info()\n",
        "# num_col\n",
        "# data.shape\n",
        "\n",
        "data = data.drop(['state_RST'], axis=1) \n",
        "\n",
        "print(data.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wPfMMkU9JPig",
        "outputId": "0de82d0a-e984-4499-f27f-9fbf9de3f252"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(33,)\n",
            "(81173, 19)\n",
            "(81173, 56)\n",
            "(81173, 53)\n",
            "(81173, 52)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Split the data set\n",
        "Y = train_set.label\n",
        "X = train_set.drop(['label', 'attack_cat'], axis=1)\n",
        "\n",
        "# transfer with label encoding\n",
        "X = X.apply(LabelEncoder().fit_transform)\n",
        "\n",
        "X_train, X_val, y_train, y_val = train_test_split(X,Y, test_size=0.2, shuffle=True,random_state=42)"
      ],
      "metadata": {
        "id": "2WFHlHglhJcT",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:06:18.808432Z",
          "iopub.execute_input": "2021-12-11T01:06:18.809205Z",
          "iopub.status.idle": "2021-12-11T01:06:18.843627Z",
          "shell.execute_reply.started": "2021-12-11T01:06:18.809163Z",
          "shell.execute_reply": "2021-12-11T01:06:18.842533Z"
        },
        "trusted": true
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "Y.value_counts()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FTo4rJk_HwGD",
        "outputId": "f8c522e1-fc71-4918-8690-bd03d728ab97"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1    61685\n",
              "0    19488\n",
              "Name: label, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 104
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Function for plot ROC curve\n",
        "from yellowbrick.classifier import ROCAUC\n",
        "\n",
        "encode_label = le3.fit_transform(Y.unique())\n",
        "dict_code = dict(zip(y.unique(),encode_label))\n",
        "\n",
        "def plot_ROC_curve(model, xtrain, ytrain, xtest, ytest):\n",
        "\n",
        "    # Creating visualization with the readable labels\n",
        "    visualizer = ROCAUC(model)\n",
        "    try:                                    \n",
        "      # Fitting to the training data first then scoring with the test data                                    \n",
        "      visualizer.fit(xtrain, ytrain)\n",
        "      visualizer.score(xtest, ytest)\n",
        "      g= visualizer.show()\n",
        "    except AttributeError:\n",
        "      pass\n",
        "    return "
      ],
      "metadata": {
        "id": "9ErgDlswVcGB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Random Forest**"
      ],
      "metadata": {
        "id": "wfodkdtLjBpW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#criterion=\"gini\",max_depth=20,min_samples_leaf=4,min_samples_split=5,random_state=42, n_jobs=-1\n",
        "rf = RandomForestClassifier(criterion=\"gini\",max_depth=20,min_samples_leaf=4,min_samples_split=5,random_state=42, n_jobs=-1)\n",
        "rf.fit(X_train,y_train).score(X_val,y_val)"
      ],
      "metadata": {
        "id": "i4HrlNJmeVuA",
        "outputId": "99a21d06-059a-4659-bbc1-cb49ca4e31e1",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:06:55.039436Z",
          "iopub.execute_input": "2021-12-11T01:06:55.040000Z",
          "iopub.status.idle": "2021-12-11T01:07:07.676248Z",
          "shell.execute_reply.started": "2021-12-11T01:06:55.039963Z",
          "shell.execute_reply": "2021-12-11T01:07:07.675408Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9868801971050201"
            ]
          },
          "metadata": {},
          "execution_count": 107
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_ROC_curve(rf, X_train, y_train,X_val,y_val)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 445
        },
        "id": "-L6nP-owWGwX",
        "outputId": "ccf2f431-f3da-4c65-e1aa-21bfa971b991"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/base.py:446: UserWarning:\n",
            "\n",
            "X does not have valid feature names, but RandomForestClassifier was fitted with feature names\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfYAAAFnCAYAAABU0WtaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeVwUdfzH8dfswgJyqKhoinepqZmK8kvNGwPNo0zzKO3wKk0tyzLzqLwzLbUyTVMjy4O8ygPNW/NAMkvNW1TUOFQOAZeFnd8f/tifBLhgLMPA5/l4+JCdnZ1573B89vud73xHUVVVRQghhBBFgkHrAEIIIYTIP1LYhRBCiCJECrsQQghRhEhhF0IIIYoQKexCCCFEESKFXQghhChCnLQOIIqn2rVrU6VKFYxGIwDp6ek0bdqUcePGUaJECQCio6OZPXs24eHhGI1GXFxc6N27N3369LFtJzU1lS+//JLQ0FAyrtwMCgpi2LBhmEymLPvN6/paWL58OV999RUvvvgir7/++gNt49ChQwwYMABfX18A23vt2bMnAwcOzLes69evJyQkhODg4HzZXrt27VBVFRcXF9syJycnfvnll3zZ/v2sWrWK559/3vZ49+7dfPXVV8TFxZGWlkatWrV45513qFmzJmvWrGHDhg0sXbo0XzMEBQXx/fffU7p0aV555RUiIyOZP38+I0eO5Pvvv6ds2bL5uj9RRKlCaKBWrVrq9evXbY/NZrM6dOhQdfbs2aqqqmpSUpL61FNPqZ9//rlqsVhUVVXVK1euqM8884w6b9482+tGjhypDh48WI2Pj1dVVVVv3bqlDh48WB01alS2+83r+lro37+/umrVqv+0jYMHD6oBAQGZlsXExKht27ZV9+3b95+2fa9169apL774Yr5tr23btmpYWFi+bS+3oqOj1Q4dOtge79y5U23RooV65MgRVVVV1Wq1qitWrFD9/f3V2NhY9aefflJfeuklh+W5du2aWqdOHTU1NdVh+xBFl3TFi0LBZDLRsmVL/v77bwDWrl2Lt7c3I0eOxMnpbseSr68v06dPZ9GiRSQmJnL27Fl2797NjBkz8PLyAqBUqVJMnTqVHj16ZNlHbtbv168f69evt73m3se1a9dmwYIFBAYGMmPGDCZNmmRb7+bNmzRs2JDExETOnTvHiy++SGBgIF26dOGvv/4CICkpiWHDhtGxY0fat2/PuHHjsFgsmTJ+8skn/PHHH8yZM4d58+ZhNpuZMGECgYGBdOzYkenTp5Oeng7cbd1+8cUXBAYGcu3aNbvHuGzZsjRs2NB2jC9cuECfPn3o2LEjHTp0yNQqrl27NuvWreOZZ57hySeftLVMrVYrH3/8MW3atKFHjx6cOnXK9pq4uDhGjhxJYGAgnTp1YuHChZm2t2rVKrp06ULr1q05cOAAo0aNom3btgwcOJC0tDS7+e1tP+N7k56enufvQe/evbl27RpBQUGkpqYyb948hg8fjp+fHwCKotCrVy9mzJiRqTcBIDY2lgEDBhAUFES7du1YsmSJ7bnvv/+ejh07EhQURI8ePTh79ux9l9euXZurV6/Sr18/rFYrXbp04dSpU9SuXZt//vkHgJUrV9r2NWrUKO7cuQPAmDFjmDZtGl26dGHz5s12j6couqSwi0IhPj6eX375hUaNGgFw+PBh2rZtm2W92rVr4+3tzZ9//snhw4dp2LAhpUqVyrROmTJlaNasWZbX5nX97KiqSmhoKB07dmTnzp225Tt37uSJJ57A3d2dYcOG0a1bN0JDQ/nwww8ZOnQoaWlprFu3Di8vLzZv3kxoaChGo5Fz585l2v67775LgwYNGD16NMOHD2fZsmX8888/bNy4kbVr13LkyJFMBTgqKorQ0FAqVqxoN/v58+c5ePCg7Rh/8skntG3bls2bNzN16lQ++OCDTB80zp07x7p16/jqq6+YPXs26enp7N27l/3797Nx40a+//57jhw5Ylt/9uzZlCxZktDQUH744Qd+/PHHTM/funWLn3/+mU6dOjFixAiGDx9OaGgoZ86cISwszG5+e9vP+N4oipLn78HUqVN56KGH2LJlC2lpaZw4cYI2bdpkydCmTRs8PDwyLZs/fz6+vr5s2bKFZcuWMWvWLK5fv87t27eZM2cOq1evZsuWLQwYMIBdu3bluDyD0Whk6dKlGI1GtmzZQp06dWzPHTlyhDlz5rBs2TJ27NiBh4cHc+bMsT1/4MABQkJC6Nixo93jKYouOccuNNOvXz+MRiMWi4X4+HhefvllBg0aBNwt9KVLl872dWXLliU+Pp74+HjKlCmT6/3ldf3sZPyxb9CgAaqqcurUKerUqcO2bdvo2LEjFy5c4MaNG7YeAD8/P7y9vTl69Kjt/3379uHv789HH31kd3+7du3i1VdfxcnJCScnJ7p06cL+/fvp1q1bpjzZuX79OkFBQQDcvn0bV1dXPvjgA1sr9KuvvrKde/fz88NsNhMTE2P7kJCxj3r16mE2m7lx4wZhYWG0bt0ad3d3gEwfcHbv3s2CBQuAuz0hHTp0YP/+/TRp0gSAgIAAAGrVqkXlypWpXr06AFWrViUqKsqWe/To0ZlaxZUrV+abb76xu/2MY/Eg34NDhw7Z9peQkICqqrn+WRk3bpytF6Vy5cqUK1eOyMhIGjZsiKIohISE0LlzZ1uxtVgs2S7PjR07dtCpUyfKly8PQJ8+fXjjjTd47733AGjWrFmWHgVR/EhhF5oJDg6mQoUK3Lx5k6CgIDp16mTrdi9dujTR0dHZvi42NhZvb2/i4+MzFQR7Spcunaf1s3Nva/+pp55i+/btVKlShd9//51PP/2UM2fOcOfOnUx/rG/fvk1cXBwdO3YkPj6eOXPmcOHCBbp27cr7779/30F7N2/epGTJkrbHJUuW5MaNG5ke5ySjBQqwd+9ePv74Yzp06GB7fu/evcyfP59bt26hKAqqqmK1Wm3Pe3p6AtgGOFqtVuLj4/Hx8bGtk3FKIyPrvY+9vLwyfQ8zPgwYDAbb1xnbv3e/M2fOtBXrfx+L+20/43uTkJCQ5+/BvUqWLInBYCAqKopKlSplyfFvf/31l62VbjAYiImJwWq14uzszNKlS/n666+ZN28etWvXZuLEidSuXTvH5fYkJiaybds29u3bB9ztpbi3l+V+Pw+i+JCueKE5b29v+vXrx8yZM23LWrVqxfbt27Ose+bMGeLj42nQoAH+/v4cO3YsS7FOSEhgzpw5ttZohtysbzAYMhWZ+Pj4HHMHBgayY8cO9u3bR9OmTfHw8MDHxwd3d3e2bNli+7dv3z5bQe3duzerV69m06ZNnDhxgnXr1t332JQtW5a4uDjb47i4uAcaGd2yZUsqVKjADz/8ANxtNb755pu8/vrrhIaGsmHDBhRFsbsdLy8vEhMTbY9v3ryZ71lzktvt/9fvgZubGw0aNGDr1q1Ztr106VIuX76cadno0aMJDAwkNDSULVu2ZOppqlu3LnPnzuXAgQM8+eSTTJw48b7L7fHx8eHZZ5+1va/Q0FD27NmTq9eK4kMKuygUXnnlFY4ePcrhw4cB6Nq1K2lpaUyfPt3WIrl27Rpjxoxh6NChlChRgpo1a9KpUydGjRpFbGwscPeP/ahRo2yt0HvlZv1y5crZBoQdPXqUiIiIHDM3atSIGzdusGbNGlvrsFKlSlSoUMHWUr558yajRo0iOTmZL7/8kpCQEADKly+Pr6+v3WLapk0bQkJCSE9PJzk5mfXr19O6deu8HFqbt956i/nz5xMfH09KSgrJycnUr18fgGXLluHs7ExycvJ9t9GoUSP27dtHSkoKKSkptveZkXXlypW2971t27b7nirIq9xu/0G+B05OTiQnJ9sG8Y0cOZKvv/7aVjRVVeWHH35g2bJltp6MDDdu3KB+/fooisLatWttx/b06dOMGDGC1NRUTCaTbZ2cludGu3bt2Lp1q+0D1a+//pppEKEQIF3xopDw8PBg8ODBzJgxg5CQEIxGI0uWLOHTTz+lY8eOODk54eLiwosvvkjPnj1tr5s0aRLz58/nhRdeQFEUnJ2d6dq1KwMGDMh2P/bWf+WVVxg1ahR79uzB39+fFi1a5JhZURQCAgJYvXo1s2bNsi2bPXs2H374IZ9//jkGg4FXXnmFEiVK0K1bN95//32++eYbFEXh8ccft53Hzkm/fv24cuUKTz/9NIqiEBQU9MADoxo3bkyjRo2YP38+Y8aMYeDAgTzzzDOUKVOG119/nYCAAF577bX7XjPetm1bdu3aRVBQEGXLlqV169a2AWxvvvkmH374IUFBQRgMBgYPHkyDBg0eKGt2crv9B/kepKamUrJkSVq0aMHatWtp3rw5s2fPZu7cuUyaNAmj0Ui9evVYvnx5lrEfI0eOZNiwYZQqVYrevXvTq1cvxo8fzw8//ICvry+dO3fG2dkZd3d3JkyYQK1atbJdnhv16tXjtddes42aL1OmTK7GaojiRVH/3V8phBBCCN2SrnghhBCiCHFoYT9z5gwBAQF8//33WZ777bff6NGjB7169eLLL790ZAwhhBCi2HBYYU9OTmbSpEk5TvwxefJk5s2bx48//sj+/fuzTNQhhBBCiLxzWGE3mUx88803ma55zXDlyhVKlizJQw89hMFgsE0xKYQQQoj/xmGj4jNmyspOTEwM3t7etsfe3t5cuXIlx21ZrVaSkpJwdnbO9WUhQgghhJ5lTEDk7u6OwZD7drguLndLSkrizJkzWscQQgghClytWrWyzJ9wP5oUdh8fH9sEIXD3RhbZddlncHZ2Bu6+ueym31RVlWtxZzkb9Qexty8SerYMgY/cyLJedlQVFAXSrQaSLC7csbijKj64uz6Eu0uZ//uUlIKBdAyKEUUxoGDAYDBiUAwoijOqquLp4oSb8/8fTkVRUBRQ7n0MmZahgML/L7+7SPnXOorta+8SJgyKcnct2/rYejH+f/tKxkuz5Mm83j37BP46cYLH6tfL9HrpIcl/x48ft00MIxxDjrHjyTHOfxlTIZvNZjZv3oybmxtnzpyx1cDc0qSw+/r6cvv2bSIjI6lQoQI7d+7k008/zXH9jOJiMpky3eAgJTWR41f3cPqf30lLT7Itb/dwMpb/uzrfqioYDV54uFXj4bK+lPOsgLtLKZyMzrg6e2A06KLTokA4GRRKuLlpHaNYkBt1OJ4cY8eTY5w/0tPTMRqNuLi48N577+Hu7o6Pjw9msxnIewPLYVXt+PHjzJgxg6tXr+Lk5ERoaCjt2rXD19eXDh068OGHH/L2228D0KlTJ9udnnJLVa3sPLWc6IQIW6v7Xh4u3nRtNAKTk2t+vSUhhBAi36iqyuzZs9myZQubNm3C2dnZ7myUueGwwl6/fn2Cg4NzfL5p06a2eZ8fRHTCZaIT7t6M4d6iXrvC/+Bfo4u0xIUQQhRqiqJw8eJFrl27xsWLF6lVq1a+bFd31c+clsz5qKMcvvizbdnVBFdGBYzGzeR+n1cKIYQQ2kpLS2P37t20b98egKlTp2K1WjPdEvq/0t2UsgfOrc1U1L8J82Vk+w+kqAshhCj0hg8fTs+ePdm7dy9w91bI+VnUQWct9mu3zhAR+5ft8fhfa7BvRF88XPI2YlAIIYTQwqBBg1AUxaFXFOiqxX7y2m+2r1f95cPeEX2p6u2hYSIhhBAiZ2fOnKF3795ER0cDd2+f/NVXX2W5/W9+0lVhj0uJsn09sdOLVJOiLoQQohDbuXMnW7duZe3atQW2T111xWeIS3Hjf6rmPKGNEEIIoZWIiAiqVKmCwWBg0KBBPProo7Rq1arA9q+rFnuGhLTHtY4ghBBCZBEaGkrz5s1ZsGABAAaDoUCLOui0sDepKtMYCiGEKHwaN25M1apVqVy5smYZdNcVb1WhSukyWscQQgghsFqtLFy4kKZNm+Ln50e5cuXYv39/nu7Glt9012K/leJErXJeWscQQggh+OOPPxg7diwTJ060LdOyqIMOW+w3kk2YnIxaxxBCCFFMWa1WzGYzbm5uNG7cmC+++IIOHTpoHctGdy12d5eyWkcQQghRTMXGxtKtWzdGjx5tW9a3b1/KlSunYarMdFfYG1eSa9eFEEJoo2TJkty+fZv4+HgsFovWcbKlu674GuVqax1BCCFEMXL58mVOnz5Nhw4dcHZ2Zt26dXh5eeX5PukFRXeFvYxHRa0jCCGEKCbMZjMdO3bk9u3bHDlyhHLlylGyZEmtY92X7gq7ouju7IEQQgidUVUVRVFwcXGxjXgvW1YfY7x0V9hNTm5aRxBCCFFEqarKsmXLWLduHSEhITg5OfH8889rHStPdNf8NUiLXQghhIMoisLhw4f5448/OH36tNZxHohUSSGEEMWaqqqEhYXZHk+dOpXffvuNevXqaZjqwemusBsU3Z09EEIIUYi99957BAYGsn//fgBKlSpFxYr6HaituyrpZpLr2IUQQuSfnj17cunSJapWrap1lHyhuxa7EEII8V9cu3aNIUOGEBsbC0DTpk1ZuXIlvr6+GifLHzor7IVzMgAhhBD6sWHDBlavXs13332ndRSH0F1XvBBCCJFXMTExlClTBoPBwKBBg6hYsSJdunTROpZD6KzFLoQQQuTN3r178ff359tvvwXAaDTStWvXQjsl7H+lq8Ku6CuuEEKIQqBWrVp4eXnh4uKidZQCIV3xQgghihRVVfnpp594+OGHadiwIeXLlycsLAyTyaR1tAIhTWAhhBBFyvHjxxk8eDDvvvsuqqoCFJuiDjprsSuKUesIQgghCiFVVUlLS8PZ2ZnHHnuMadOmERgYWGTPo9+Pvgq7dDAIIYT4l/j4eIYPH06pUqWYO3cuAEOGDNE4lXZ0VdhV0rWOIIQQopBxc3MjIiICT09P7ty5g6urq9aRNKWrwm5V07SOIIQQohCIjY3l9OnTtGjRApPJREhICGXLlsVgkJ5dXRV2o1J8Bj8IIYTInsVi4amnnuLWrVscPHiQ8uXL4+Pjo3WsQkNXhV2mlBVCCOHs7MyoUaOIj4+nbNmyWscpdHRV2GVUvBBCFE+//PILP/74I8uWLcPJyYkXX3xR60iFlq5ORijSYhdCiGJpw4YN7Nixg2PHjmkdpdDTVWEXQghRfJw6dcr29fTp09m1axd+fn4aJtIHfRV2RV9xhRBCPJjJkyfTokULDh48CIC3tze1a9fWOJU+6KpSWq2pWkcQQghRAAICAmjcuDGlS5fWOoru6Kqwu7nI6EchhCiK4uPjGTNmDDdv3gTgiSeeYOvWrdJKfwC6KuzSYhdCiKJpxYoVLFy4kPnz59uWFcd53vODri53U1WZUlYIIYqKhIQEPD09URSFgQMH4ubmRp8+fbSOpXu6arGnpZu1jiCEECIfhIeH07x5c5YuXQqA0Wikf//+ODs7axusCNBVYXc1ldE6ghBCiHxQsWJFLBYLiYmJWkcpcnTVFW+QmeeEEEK3duzYgY+PD/Xr1+ehhx4iPDwcDw8PrWMVObpqsQshhNCnU6dO0aNHD0aOHImqqgBS1B1EVy12qwyeE0IIXbFarRgMBurUqcO4ceMICAiQ0e4OpqvC7mRw0TqCEEKIXEhOTmbChAkAfPrppwCMGjVKy0jFhq4KuxBCCH0wGo0cOHAARVFITk6mRIkSWkcqNnRV2NOtFq0jCCGEyMHt27c5ffo0fn5+uLi4sGLFCnx8fHBxkd7WgqSrwm5VZeY5IYQojNLT0wkKCuLatWscOHCA8uXLU7lyZa1jFUsOLexTp07l2LFjKIrC2LFjadCgge255cuXs2HDBgwGA/Xr1+eDDz5wZBQhhBAOZDQaefXVV7ly5QolS5bUOk6x5rDCfvjwYS5dusTKlSs5f/48Y8eOZeXKlcDd7prFixezdetWnJycePXVV/njjz9o2LDh/Tcq17ELIUSh8dtvv/Hdd9/x5Zdf2gq70J7DrmM/cOAAAQEBANSsWZP4+Hhu374NgLOzM87OziQnJ5OWlkZKSkquPuEZFZlqUAghCovFixcTEhLCoUOHtI4i7uGwwh4bG5vpPrre3t7ExMQA4OLiwrBhwwgICKBt27Y8/vjjVK9e3VFRhBBC5JPIyEjb15988gmbN2+mefPmGiYS/1Zgg+cyZhqCu13xCxYsYMuWLXh4ePDSSy9x6tQp6tSpc99tJN6OIzw83NFRizU5vgVDjrPjyTHOf6tXr2bx4sXMnj2bunXrEhERgcFgkGNdyDissPv4+BAbG2t7HB0dTbly5QA4f/48lStXxtvbG4AmTZpw/Phxu4Xdy6MMfo39HBW52AsPD8fPT46vo8lxdjw5xo5hNpvZuXMndevWBZBj7GBms5njx4/n+XUO64pv0aIFoaGhAJw4cQIfHx/bvMCVKlXi/Pnz3LlzB4Djx49TrVo1u9t0cnJ1VFwhhBD/kpKSwvTp04mLiwOgefPm7N+/Xwp6IeewFnvjxo2pV68evXv3RlEUJk6cyJo1a/D09KRDhw4MGDCA/v37YzQaadSoEU2aNLG7TWejzFwkhBAFZfny5XzyySckJSUxadIkAJycdDX9SbHk0O/QO++8k+nxvV3tvXv3pnfv3nnaXrrVnC+5hBBCZM9sNmMymVAUhZdffpnk5GQGDBigdSyRBzq7bavcEUgIIRzl5MmTtG7dmuDgYOBu63zEiBG4u7trnEzkhRR2IYQQAJQsWZKoqCgiIiK0jiL+A12dLDE5eWodQQghipSjR4/i5uZGnTp1qFSpEkeOHKFMmTJaxxL/gc5a7EIIIfLLhQsXeOqppxg6dChWqxVAinoRoKsWu1VN0zqCEEIUGTVq1GDkyJG0bNkSg0HaeUWFrgq7osgPnhBCPKjU1FRmzZpFYmIiU6dOBWDcuHEapxL5TVeV0mgwaR1BCCF0y2q1smHDBjZu3EhiYqLWcYSD6KvFrnUAIYTQGYvFwtmzZ6lbty6urq4EBwfj4+ODp6cMRi6qdNViN1vitY4ghBC6YbVa6dKlC926dbPdXfPhhx/Gy8tL42TCkXTVYldR7a8khBACAIPBwDPPPMPJkydxcXHROo4oILoq7EIIIe7v5MmTLFmyhBkzZmAwGBgyZAiKIicyixNddcU7G2VaQyGEuJ9Zs2axePFidu7cCSBFvRjSVWFXDEatIwghRKFz8+ZN29fTpk1jxYoVtG/fXsNEQkv6KuxaBxBCiEJm2bJlNGjQgN9//x0AHx8fnnrqKY1TCS3pqrCnpadqHUEIIQqVGjVq4OXlJdelCxtdFXaTUwmtIwghhKbS09NZsGAB8fF3L/9t2bIl4eHhtG7dWuNkorDQVWEXQojiLjg4mPfff59p06bZlrm5uWmYSBQ2urrcLTUtWesIQghR4NLT0zEYDCiKwgsvvEBkZCRDhw7VOpYopHTVYletco5dCFG8RERE0KlTJ3788UcAnJ2dGTduHN7e3honE4WVrgq7EEIUN0ajkb///pvDhw9rHUXohK664g1GmRJRCFH0XbhwAavVysMPP0zlypXZv38/lStX1jqW0AldtdidDFLYhRBF25UrV2jZsiVDhgwhPT0dQIq6yBNdtdiFEKKoq1y5Mv3798ff3x+jUWbbFHmnq8Kenm7WOoIQQuQrq9XK4sWLiYqKYty4cQCZLmUTIq901RXv7CTXagohipbU1FQWLVrEsmXLiIuL0zqOKAJ01WIXQoiiwGq1cvnyZapVq4arqytLliyhTJkylCpVSutoogjQVYs93WrROoIQQvwnqqrSt29fAgMDuXHjBgB169alfPnyGicTRYWuWuxp6SlaRxBCiP9EURSefPJJDAaDbdS7EPlJVy12IYTQoytXrvDxxx9jtVoBGDp0KMuXL8fHx0fjZKIo0lVhVxRddTAIIQQAH3/8MZ9//jlbtmwBsM37LoQj6KpSyqh4IYReJCUl4e7uDsCkSZNo164dHTt21DiVKA701WJHPuEKIQq/tWvX8vjjj3Ps2DEAKlSoQJ8+faSVLgqErgp7utzdTQihA6VKlSItLY3IyEito4hiSFeF3Whw1jqCEEJkoaoqq1evJiEhAYC2bdty7Ngxnn76aY2TieJIV4Vdd3GFEMXCypUrGTJkCJMnT7YtK1mypIaJRHGmq8FzYNU6gBBCAHdb6XD3uvTu3bsTHh7O8OHDNU4lhM6awKlpSVpHEEIIoqKieOGFF1i9ejUAJpOJmTNnyu1VRaGgq8IuhBCFwZ07d9i7dy8bN27UOooQWeirK16uFBFCaCQqKoo7d+5QtWpVqlatytatW6ldu7bWsYTIQlctdpPRU+sIQohi6Pr16zRv3pxBgwbZ5nd/9NFHMRh09SdUFBO6arHL5A5CCC089NBDdO7cmfr168vfIVHo2f24efXqVUaMGEG/fv0AWLVqFREREY7OlS1VlTshCSEKxtq1a/n0009tj+fMmcOgQYOklS4KPbs/oePHj6dbt262SzuqV6/O+PHjHR4sOyqqJvsVQhQvZrOZKVOmMHfuXGJjY7WOI0Se2C3sFouF9u3b27qfmjZt6vBQOTEoMvOcEMJxoqKiAHBxcWHRokXs3LmTsmXLapxKiLzJVZ9SQkKCrbCfPXsWs9ns0FBCCFGQVFVl6NChtGnThlu3bgHQsGFDatasqXEyIfLO7uC5YcOG8fzzzxMTE0OXLl24desWM2fOLIhsWVjSZYIaIUT+UxSF2rVrc+HCBW7fvk3p0qW1jiTEA7Nb2OvWrcu6des4c+YMJpOJ6tWrEx0dXRDZssg4zy+EEP/VzZs3+e677xg5ciSKovDGG2/wxhtvYDQatY4mxH9y3654q9XKsGHDcHFxoX79+tSqVQtFURg6dGhB5RNCCIf44IMP+Pjjj1m/fj0ARqNRirooEnJssf/yyy/MmzePS5cu8eijj6IoCqqqYjAYePLJJwsyo42z0VWT/QohigaLxYKz891BuBMmTKB+/fp06dJF41RC5K8cW+ydO3cmNDSUYcOGcerUKf7++29OnTrFyZMnmTVrVkFmtJFR8UKIB7V9+3b8/Pz466+/gLuTzgwbNkxa6aLIsTsqfvjw4Zw7d46wsDDCwsLYv38/zz//fEFky0LmexJCPCir1UpMTAwnTpzQOooQDklXVGQAACAASURBVGV38NyUKVPYt28fsbGxVKlShStXrvDqq6/mauNTp07l2LFjKIrC2LFjadCgge2569evM2rUKCwWC3Xr1uXjjz+2u700651c7VcIIeBuK93f3x9PT086dOjA0aNHqVChgtaxhHAouy32P//8k82bN1OnTh1++uknvv32W1JSUuxu+PDhw1y6dImVK1cyZcoUpkyZkun56dOn8+qrrxISEoLRaOTatWt2t2k0uNhdRwghADZs2EDPnj2ZNGmSbZkUdVEc2C3sJpMJuDvoRFVV6tevz++//253wwcOHCAgIACAmjVrEh8fz+3bt4G7XWLh4eG0a9cOgIkTJ1KxYsUHfhNCCPFvgYGB9OzZk5dfflnrKEIUKLtd8dWrV2f58uU0adKEV155herVq5OYmGh3w7GxsdSrV8/22Nvbm5iYGDw8PLh58ybu7u5MmzaNEydO0KRJE95++22724y8epnwBLkBgyOFh4drHaFYkOOc/5KSkvj6669p3Lgxbdu25fjx4wwePJiUlBQ53g4ix7VwslvYP/roI+Lj4/Hy8mLjxo3cuHGDIUOG5HlH904uo6oqUVFR9O/fn0qVKjF48GB27dpFmzZt7ruN8hXK4FfHL8/7FrkTHh6On58cX0eT4+wYFy9eZM+ePaSkpNCmTRuaNGmidaQiTX6OHc9sNnP8+PE8v+6+zd+EhAROnDiBi4sLBoOBLl268PLLL9tulHA/Pj4+me6KFB0dTbly5QAoXbo0FStWpEqVKhiNRpo1a8bZs2fzHF4IUbwlJCRw9epV4G7v4vr161m1apXcM10UazkW9m3bttGpUyfGjx9Phw4dOH78OKmpqcyYMYN33nnH7oZbtGhBaGgoACdOnMDHxwcPDw8AnJycqFy5su2+7idOnKB69ep2t2lUTLl5T0KIYiA2NpYWLVowcOBA0tPTAfDz87NNQCNEcZVjV/zixYtZv349ZcqU4fjx40yYMAGz2cyTTz5pm4Lxfho3bky9evXo3bs3iqIwceJE1qxZY7vsZOzYsYwZMwZVValVq5ZtIN19wzrJzHNCiLvKli3LE088QY0aNbBarTLRjBD/J8fC7uzsTJkyZQCoX78+d+7cYcaMGTz22GO53vi/W/Z16tSxfV21alV+/PHHvOYVQhRje/bs4a+//mLYsGEALFy4ULrdhfiXHAv7v39ZypQpk6ei7ghp6TJBjRDFlcVi4a233uLq1as899xzVKhQQYq6ENnIsbCrqmr79+9lAAZDwV92JhPUCFH8JCQk4OXlhbOzM19//TVOTk4y0YwQ95FjYQ8LC6Nu3bq2x6qqUrduXVRVRVEU/v777wIJKIQovsaNG8fatWvZv38/pUqVomnTplpHEqLQy7Gwnzp1qiBz5IpVTdc6ghCiAJUuXRpPT09iYmIoVaqU1nGE0AVdTeOWLjeBEaJIS0pK4ttvv7Wd8hs5ciS7du3ikUce0TiZEPphd+Y5IYQoKGPHjiU4OJiSJUvy3HPP4eTkhJOT/JkSIi909RujKLrqYBBC5ILVarUNxh09ejRly5alU6dOGqcSQr/sVsrU1FSWL1/Op59+CsCxY8cwm80OD5Ydk9Fdk/0KIRwjLCyMli1bcvLkSQB8fX0ZP348bm5uGicTQr/sFvYPP/yQy5cvc+jQIeDu9K9jxoxxeLDsKMg1q0IUJTdv3uTUqVPs379f6yhCFBl2C/uFCxd4//33cXW9O51r3759iY6Odniw7KSrFk32K4TIP0eOHCEpKQm4e8/0w4cPM2jQII1TCVF02C3sGQNXMmZ4Sk5O5s4dbUanq6q02IXQs23bthEUFMSkSZNsy2rWrKlhIiGKHruD54KCgnjppZeIjIxk8uTJ7Nmzh759+xZEtiwMMnhOCF1r2bIlHTp0oFu3blpHEaLIslvYX3zxRRo0aMDhw4cxmUzMnj2b+vXrF0Q2IYTOZdw8qmHDhnTr1g1XV1e5+ZMQDma3sD///PN069aNHj16aD7zkyU9RdP9CyHy5tq1ayxYsIB69erRtWtXuWmLEAXAbt/2e++9x8WLF3n22Wd5/fXX2bJlC6mpqQWRLQv5kyBE4Wc2m20DbGvUqMGKFStYu3atFHUhCojdwu7n58e4cePYsWMHL7/8Mnv37qVVq1YFkU0IoTNxcXG0a9eOV199FavVCkCrVq3w8PDQOJkQxUeuZp5LSEjg119/ZcuWLVy5coVevXo5Ole2nJ1kghohCrOSJUtSs2ZNypUrR2pqqu0yWSFEwbFb2AcMGMDZs2cJCAjgtddeo3HjxgWRK1sGxajZvoUQ2fvzzz8JDw/nlVdeQVEUvv32W5nfXQgN2f3t69+/Py1btrTN5awtq9YBhBD3SE9PZ8CAAVy6dIkOHTrg6+srRV0IjeX4Gzh58mTGjRvHggULWLhwYZbnly9f7tBg2UlL12bQnhAiszt37uDq6orRaGTOnDncuXMHX19frWMJIbhPYe/RowcAb775ZoGFscdokJaAEFr77LPPCA4OZteuXXh5edG8eXOtIwkh7pFjpaxTpw4Aa9asYfr06ZmeGzBgAP7+/o5NJoQolFJTU0lNTeXSpUs89thjWscRQvxLjoV9w4YNrFixgrNnz/LCCy/YllssFm7cuFEg4f5NuuKFKHgWi4UNGzbQvXt3FEXhrbfe4rXXXqNkyZJaRxNCZCPHwt61a1f+53/+h3feeYfhw4fblhsMBh5++OECCfdvqgyeE6LAjR8/noULF2K1WunZsycmkwmTyaR1LCFEDnIs7NHR0ZQvX56pU6dmeS4xMVHz6WWFEAVj2LBhmM1mOnTooHUUIUQu5FjYZ8yYwaxZs3jppZdQFAVVVW3PKYrC9u3bCyTgvZyMLgW+TyGKm1OnTvH2228za9Ys6tSpQ+XKlfnss8+0jiWEyKUcC/usWbMA2LFjR4GFsUcKuxCOd+HCBQ4cOMDGjRttg2iFEPphd9aZ3bt3s379egDefvttnnrqKbZu3erwYEKIgnPmzBmSk5MB6NSpE9u3b+ftt9/WOJUQ4kHYLexfffUVLVu2ZPfu3VitVtauXUtwcHBBZMvCknZHk/0KUZTt27eP1q1bM2XKFNuyRo0aaZhICPFf2C3srq6ueHt7s3v3brp164a7u7tm08saDTJXvBD5zc/Pj8aNG9OsWTOtowgh8oHdCm02m1m0aBF79uyhWbNmREREkJiYWBDZhBAOkJ6ezpdffsmmTZsAcHNz45dffqFz584aJxNC5Ae7hX3SpElERUUxffp0XFxc2LdvH++8805BZMvCak3XZL9CFCWXL19m8uTJTJ061XbPdEVRNE4lhMgvdidff+SRR3jppZc4efIk27Zto127dlSsWLEgsmVhVdM02a8Qeme1WklISKBUqVJUr16dpUuX4ufnV0ju2iiEyE92f6t//PFH+vfvz8aNG/n555/p168fa9euLYhsQoh8kJCQQOfOnXn55ZdtLfTAwEDKli2rcTIhhCPYbbGvX7+ezZs34+Jy9xry5ORkXnnlFZ599lmHh/s3ububEHnn6elJyZIlMZlMpKSk4O7urnUkIYQD2a2UTk5OtqIOUKJECZydnR0aKucsbprsVwi9iYiI4NChQ/Tq1QtFUfj2229xdXWVc+lCFAN2C3uFChWYNGmS7Z7L+/bt46GHHnJ4sOzJHyUh7LFarfTu3ZsLFy7g7+9P9erVcXOTD8VCFBd2C/ukSZMIDg5mzZo1KIrC448/Tr9+/QoiWxbpVrMm+xVCD9LT0zEajRgMBmbMmEFsbCzVqlXTOpYQooDZLexms5nBgwcXRBb7VPurCFEcLV26lIULF7Jlyxa8vLxo3bq11pGEEBrJcVT8kSNHePLJJwkMDOTpp5/m8uXLBZkre3J+UIhsXb16levXr3Pq1CmtowghNJZjYf/ss89YsmQJhw4dYty4cba7vQkhtKeqKtu2bbPdTnn06NEcOHAAf39/jZMJIbSWY2E3GAw88sgjADRr1oybN28WWKicSHtdiLsmT55Mr169CAkJAcBkMlGhQgWNUwkhCoMcz7H/+7IYuUxGiMKjX79+nD9/nhYtWmgdRQhRyORY2OPj4zlw4IDtcUJCQqbH2twJSj5ciOIpMjKSMWPGMHHiRB555BGqVavG0qVLtY4lhCiEcizsXl5efPXVV7bHnp6etseKomhS2F2cPQt8n0IUBkePHmXTpk08/PDDfPjhh1rHEUIUYjkW9uDg4ILMIYT4l2vXruHt7Y2rqytdunRh7dq1tGrVSutYQohCTle3drJaLVpHEKJAhIWF0bx5c6ZNm2Zb1rp1axnrIoSwS1eFPS39jtYRhCgQdevWpVq1atSsWVPrKEIInZHbpQlRCKiqyqpVq/D29qZDhw64u7uzY8cOuV+6ECLP7P7VuHr1KiNGjLDND79q1SoiIiIcnUuIYuXKlSuMHDmSsWPHkp6eDiBFXQjxQOz+5Rg/fjzdunWzzXBVvXp1xo8f7/BgQhR1qqqSlJQEQJUqVfjqq6/46aefMBqNGicTQuiZ3cJusVho3769bdBO06ZNc73xqVOn0qtXL3r37s2ff/6Z7TqzZs3S7G5xQmglOTmZ/v37069fP9uH5u7du1OlShWNkwkh9C5XfX0JCQm2wn727FnMZvu3Tz18+DCXLl1i5cqVTJkyhSlTpmRZ59y5c4SFheUxshD65+bmhtlsxmKxkJiYqHUcIUQRYnfw3LBhw3j++eeJiYmhS5cu3Lp1i5kzZ9rd8IEDBwgICACgZs2axMfHc/v2bTw8PGzrTJ8+nbfeeosvvvgiV2GdjSVytZ4QhVFMTAwHDx6kYsWKKIrCokWL8PDwkHPpQoh8ZbewP/HEE6xbt44zZ85gMpmoXr06Li4udjccGxtLvXr1bI+9vb2JiYmxFfY1a9bg7+9PpUqVch024tJlzDHxuV5f5F14eLjWEYokVVV5/fXXuXTpEgsXLtQ6TrEgP8uOJ8e4cLJb2OfMmZPt8pEjR+ZpRxnnEQHi4uJYs2YNS5YsISoqKtfbqFG9BnV9q+VpvyL3wsPD8fPz0zpGkaKqqu001rRp0zh37hyVKlWS4+xg8rPseHKMHc9sNnP8+PE8v85uH6DRaLT9s1qtHDp0KFfnBH18fIiNjbU9jo6Oply5cgAcPHiQmzdv8sILL/DGG29w4sQJpk6daneb5jQ5Fyn0Y8OGDTz11FPcvn0bgICAAF577TXpehdCOJTdFvsbb7yR6XF6ejrDhw+3u+EWLVowb948evfuzYkTJ/Dx8bF1wwcFBREUFATcvWvV+++/z9ixY+1uU2bTFHpy7NgxTpw4QXh4OK1bt9Y6jhCimMjzzHNpaWlcvnzZ7nqNGzemXr169O7dG0VRmDhxImvWrMHT05MOHTo8UFghCruwsDCaNGmCoii8++679O3bV6aFFUIUKLuF/d83noiPj+fZZ5/N1cbfeeedTI/r1KmTZR1fX1+5k5woEj777DMmTZrEokWL6N69Oy4uLlLUhRAFzm5h/+GHH2xfK4qCh4cHXl5eDg0lhB5169aN3bt3Z7oaRAghCprdUTwzZ86kUqVKVKpUiYoVK0pRF+L/3Lp1i+HDh3PhwgUAatSowbp166hdu7bGyYQQxZndFruvry8hISE0atQIk8lkW165cmWHBsuOUbF//bwQBWXPnj0sX74cV1fXXE3aJIQQBcFuYd+0aVOWZYqisH37docEuh8noxR2oa24uDhcXV1xdXWla9euLF26lE6dOmkdSwghbHIs7Bs2bKBr167s2LGjIPMIUWgdP36cXr160atXLyZMmICiKHTt2lXrWEIIkUmO59hDQkIKMkeuWNJStI4girFq1arJ4FEhRKGX5+vYtWQlTesIopj59ddfcXJyok2bNnh4eLB3795MY02EEKKwybGwHz16lDZt2mRZnjH39a5duxwYK3sy8ZwoSNeuXeOFF16gYsWKhIWF4eTkJEVdCFHo5VjY69aty+zZswsyixCFgsViwdnZmYoVKzJ79mwef/xxnJx01bklhCjGcvxrZTKZ8nRLVSH0zmw2895773H9+nVWrFiBoii88MILWscSQog8yXHwXIMGDQoyhxCaM5lMXLlyhatXr3Lr1i2t4wghxAPJscU+evTogsyRKwaMWkcQRUxiYiKHDx+mffv2KIrCwoUL8fT0lHPpQgjd0tWNoZ2dSmgdQRQhqqrSo0cPXnjhBU6fPg1AmTJlpKgLIXRNRgSJYktRFN566y3Cw8OpVq2a1nGEECJf6KrFnpZu0TqC0Ll9+/bx3HPPkZSUBEBQUBAffPABLi4yXbEQomjQVWFPV81aRxA6t23bNnbv3s2+ffu0jiKEEA6hq8IuE9SIB3Hq1Cnb1++//z6//vorgYGBGiYSQgjH0VVhFyKvFi1aRIsWLVi/fj0Arq6uNGzYUONUQgjhOFLYRZHWunVr6tevT8WKFbWOIoQQBUIKuyhSkpOTmThxIhEREQA88sgj7Nq1i6ZNm2obTAghCoiuCrsiZ9mFHb/++ivz5s1j5syZtmWKIj83QojiQ1fXsZuc3LWOIAqhlJQUjEYjJpOJLl26MGfOHJ577jmtYwkhhCb01WKXlpf4l3PnztGmTRtbC11RFPr160eJEjJLoRCieNJVYbeqVq0jiEKmfPnypKamcufOHa2jCCFEoaCrrnhLWrLWEUQhEB4ejtlspnnz5nh6erJ37148PDy0jiWEEIWCrgq7ENHR0XTu3Jly5coRFhaGi4uLFHUhhLiHvgq7nGMvtqxWKwaDAR8fHyZNmkTt2rVlfnchhMiGvgq7KHbS0tKYMWMGp0+fZtmyZSiKwsCBA7WOJYQQhZauBs+J4sdgMBAWFsaxY8eIjo7WOo4QQhR60mIXhU5qaiq///47TzzxBAaDga+//hp3d3c8PT21jiaEEIWerlrszkZXrSOIAtCnTx+effZZTp8+DUCFChWkqAshRC7pqsVuMOgqrnhAr776KpUrV+ahhx7SOooQQuiOrlrsomg6fvw4L7/8MikpKQA8/fTTfP7553h5eWmcTAgh9EdXhT1VJqgpklasWMGGDRvYsmWL1lGEEEL39NW3rapaJxD5JDIyEl9fXwDGjh1LQEAAbdq00TaUEEIUAbpqsYuiYeXKlfj5+bFx40YASpQoIUVdCCHyib4Ku8w8VyQ0bNiQSpUq4ebmpnUUIYQocvRV2IUupaWlMXfuXK5cuQJA7dq1OXz4MO3atdM4mRBCFD1S2IXDhYaG8uGHH/LRRx/Zljk56Wt4hxBC6IWu/roa5Tp23UhPT8dqteLs7EynTp2YPHkyffr00TqWEEIUebpqsTsZ5W5eehAZGUlQUBCzZ88GQFEUhg4dSunSpTVOJoQQRZ+uCrvQBy8vL65fv87ly5dR5RJFIYQoULrq205LN2sdQeTg3Llz3Lp1i6ZNm+Ll5cWuXbsoW7as1rGEEKLY0VVht1rTtI4gsnHr1i3atWtH6dKlOXjwIG5ublLUhRBCI7oq7KJwKl26NKNHj6ZKlSpybboQQmhMCrvIM6vVysKFCzl69Chff/01iqIwfPhwrWMJDaSlpWG1WrWOka3U1FStIxR5cozzh8FgyNdLgHU1eE4mnis8Nm7cyPbt27l69arWUYRGEhMTC+0f9po1a2odociTY5x/UlNTSUxMzLftSYtd5IrVauXEiRM89thjGAwG5s+fj8lkwsfHR+toQgNpaWkYjUZKlCihdZRsWSwWTCaT1jGKNDnG+cdkMpGcnExaWlq+tNx11WIHabJrZcCAAQQGBnLu3DkAfH19pagXY1arVWYPFCIfGY3GfDutpavfTJNT4WwdFAfPPPMMFosFLy8vraMIIUSRo+TjuWaHttinTp1Kr1696N27N3/++Wem5w4ePMjzzz9P7969ef/99wvtAJzi6vLly4wcOZI7d+4A0K1bN4KDg6WVLoQQhZzDCvvhw4e5dOkSK1euZMqUKUyZMiXT8xMmTGDu3LmsWLGCpKQk9u7da3ebVjXdUXHFvyxcuJDg4GB++ukn27L8/EQpxH8VGRlJo0aN6NevH/369aNXr16MHz+e9PS7fydSUlKYMGECzzzzDD169OC1117j+vXrttdHREQwePBgevToQffu3Zk0aVKuBwP+/PPPBAYGcuTIkfuud+jQIUaMGPHgbzIb169fp1+/fvTt25eRI0fazRwVFcWjjz7Kr7/+alu2Zs0aZsyYkWm9MWPGsHPnTgBiY2N588036d69Oz169ODtt98mISHhgfJu3ryZRo0acebMmWyfX7RoET169KBnz57s3r0buDswc/DgwfTp04cBAwYQFxf3QPsurhxW2A8cOEBAQABwd/RkfHw8t2/ftj2/Zs0aKlSoAIC3tze3bt2yu02Zec6x7v3Fff/991m8eDF9+/bVMJEQ91e9enWCg4MJDg5m5cqVWCwWfv75ZwCmTZuGj48P69atIyQkhEGDBjFw4EAsFgvp6ekMHz6cgQMHEhISYvsA++WXX+Zqv7/99hujR4+mSZMmDntvOZk7dy59+/blhx9+oGrVqoSEhNx3/Y0bN1K1alU2btyY6328++67tG/fnjVr1hASEsKjjz6a6e6MuXX48GH27NlD7dq1s33+ypUrbNq0iR9++IEFCxYwbdo00tPTWbZsGf7+/vz444889dRTfPPNN3ned3HmsHPssbGx1KtXz/bY29ubmJgYPDw8AGz/R0dHs3//fkaOHOmoKCIXNm7cyJAhQ1i8eDGBgYG4u7vz7LPPah1L6MS7P4cTcuxSvm6zx+NV+aSLX55e06BBAy5dumTrBdy2bZvtOT8/Pxo0aMD27dspUaIENWrUwN/fH7jbGzV69GgMhsxtHYvFwoQJE7hy5QqpqamMGDECRVHYs2cPx48fx8vLy7YNgMmTJ/Pnn39iNBqzFMJvv/2W0NBQrFYrrVu35o033uDkyZN89NFHmEwmTCYTn332GZGRkVmW3Tu25dChQ7Ztt23blm+//fa+H8B/+eUXJkyYwFtvvUVycrLdKxnOnz9PQkICXbp0sS175ZVXbKflMuzdu5fly5dnWvb8889nel3dunXx9/enX79+2e7r0KFDtGzZEpPJhLe3N5UqVeLcuXMcOHCAqVOn2t7ja6+9dt/MIrMCGzyX3c1Abty4wWuvvcbEiRNzdeeviEsXSYnKv2v9xP9LSUnBZDJx4sQJmQ62AISHh2sd4T+rWbMmFosFuFsA8/uGPxaLhaSkpByfT0lJwWq12taxWCxs3bqVHj16EBkZSZUqVTCbzZjN/9/TV7NmTU6fPo2bmxs1a9bMsv309HTbe4K7RdFgMLBgwQJiYmIYNGgQ69ato1mzZrRv35569erZtnHo0CEiIyNZsmQJ4eHhrF+/Hn9/f9LS0khKSiI1NZVvvvkGg8FAly5d6NmzJytXrqR79+507tyZw4cPc/nyZVavXp1lWfXq1W2ZkpOTsVgsWCwW3Nzc+Oeff3I8ThEREcTHx/P444/j5+fH5s2bCQoKwmw2Zzm+aWlp3Llzh7///puHH344223eu6xly5a0bNnyvusoikJSUhLp6emkpKRk2ea1a9fw8PCwLffy8uLKlStER0fj4uJCUlISrq6uREVF3fdnoSiwWCycP38+X7blsMLu4+NDbGys7XF0dDTlypWzPb59+zaDBg3izTff5Mknn8zVNqtVrU5d31r5nrU4UlWV5cuX06ZNG3x9ffHz86NixYo0b95c62hFXnh4OH5+eWuJFjYZ53UzrmP+rPsTfNa9YDO4ublx6dIlW2vu9OnTDBw4kM6dO3P06FEURcHd3T3Ta5ydnXFzc8PFxQWLxZLl+X87e/YsLVq0wN3dHXd3d1xdXbFYLDg5OeHq6prp9efPn8ff3x93d3datWpFq1atOHToEE5OTri7u+Pl5cWQIUNwcnIiLi4Oi8VCUFAQH374IdevX6dTp07UqlWLxMTELMvude/7cnNzw2g05vg+tm/fTufOnXF3d+eZZ55hzZo1PPfcc7i4uODs7JzpdU5OTri5uZGWlobBYLB7bJKSkuyuk8FoNOLm5pZl/YxeiYzlRqMRV1dX2/7d3d1JS0vL9ntZ1KSmpvLYY49lmhvAbDZz/PjxPG/LYefYW7RoQWhoKAAnTpzAx8fH1v0OMH36dF566SVatWqV623K0K38s23bNkaMGMEHH3xgW+biIve7F/py7zn2J554wtayrVSpEhcvXswysOzUqVPUrFmTGjVq8Ndff2V6LjU1NdsBXvf2RKSmpmbprs9wv+uQr169ytKlS1m0aBHBwcFUqlQJgGbNmhESEkKNGjUYM2YMBw8ezHbZvUqUKGHrFo+KirrvlSobN24kNDSUbt26MW/ePA4ePEhCQgLe3t5ZBsPdvHmTcuXKZXtsgCwFZu/evbaBixn/MsY35Na/G4AZ78fHx4eYmJhcvUeRlcMKe+PGjalXrx69e/dm8uTJTJw4kTVr1rBt2zZSUlJsA1oyfiBWrlzpqCji/6iqahsx3KFDB959990sVysIoVejR4/m008/JSUlBXd3d9q2bcsXX3xhe/7333/n5MmTtGnThhYtWnD16lV27NgB3J1wZ+bMmWzatCnTNh977DEOHToE3B2NbjAYcpzL4d51M86dZ7h16xbe3t64u7tz4sQJrl69isVi4fvvvycuLo6uXbvy0ksv8ffff2e77F7Nmze3NZq2bt2abXc4wJ9//om7uztbtmxh/fr1/Pzzz3Ts2JHQ0FAaNGhAeHg4N2/eBO522UdGRvLII49Qo0YNKlSokOn8+ZIlS1i2bFmm7bds2dL2oSrj373n13PjiSeeYNeuXaSmphIVFUV0dDQPP/wwLVq0YMuWLXbfo8ieoub3iTEHyOiO8KrkwyMVKmsdR5eio6MZPnw4/v7+vP3229muUxS6iPWgKBznf3fFkavC1AAAIABJREFUayEyMpIRI0awZs0a27LZs2cDMGTIEJydnZk1axYHDhywDc4aO3Ys1apVA+7+TkyYMIHo6GhMJhPNmzfnjTfeyNQiT0tLY+LEiVy+fBmLxcLbb79N06ZNGTNmDIGBgbRt2zZTpunTp9vm7Jg4cSJxcXEsX76czz77jMGDB5OUlISfnx9Wq5W///6bV199lc8//xxPT09MJhPTpk3j5MmTWZbdO+4lOjqa9957D7PZTMWKFZk2bRrOzs689dZbTJs2DVdXV+DuPCJVqlThxRdftL32yJEjfPnllyxZsoT9+/fz5ZdfYjQacXJy4p133rENeL59+zYff/wxp0+fpkSJEtSpU4cxY8Zk6tXLTVf86tWr2bBhA3///TdVq1alZs2afPLJJ0yZMoX+/ftTuXJlgoOD+fnnn1EUhTfffJNmzZqRlJTE6NGjiYuLw8vLi5kzZ+Lp6Znrnw09yu53KqP21a9fP089qroq7KUqPUTNChW1jqNLcXFxtGjRgscee4wff/wx22vSi0LB0YOicJwLQ2G/n7yc/y0qZs+ezYgRIwpsqt/ieIwdKT8Lu66mlFUp9J9BCpXr168TExNDgwYNKFWqFKGhoVSqVEkmmhGiCGrYsKHM3y8AnRV2S1qy1hF0IyEhgVatWuHh4cG+fftwd3fH19dX61hCCAdp166d1hFEIaGrwi5yz8vLi8GDB1OmTBnc3Ny0jiOEEKKASGEvIlRV5aeffmL//v189tlnwN1RwkIIIYoXnd2PXeREVVUWL17M6tWriYiI0DqOEEIIjeiqxS5jvjJTVZWIiAiqV6+OwWBg/vz5qKpqu5xHCCFE8SMtdh0bNWoULVu25OLFiwBUq1Yt05zSQhRlxfW2rQDfffddpnnq70du21r86KqwG42F85pZrTz55JM0aNAgxykuhSjqiuNtW9etW8eNGzdyPc2q3La1+NFVV7yTUrwLe2xsLHPnzuWDDz7AxcWF7t278+yzz0phF5oLu7iJiNg/83Wb1co2oGn1Tnl6TXG4bWtAQAAeHh65npddbtta/OiqsBd3c+fO5YsvvqBq1aoMGDAARVFkshkh/o/FYmH79u306dOHyMhIatSokWXClkcffZSLFy/i5ubGo48+mum5jKlY77Vx40ZMJhPff/89UVFR9O/fn9DQUFq2bElgYGCmov7bb7/xzz//sGrVKsLCwti0aRPNmjXLtL0ffvgBg8FA+/btefnll1mzZg19+vThmWee4cCBA8TExGS77N7Cfu/NtOy5cOECiYmJ/G979x6X4/0/cPx1V/ctkXJIOWQOG6Z+jkM5R1nCxr5R6SSHkcjMY+mAjBznHDNz2ObMrDlksn0tfJGMmhxmtjJLTEmhk+5y//7o0TX3Opvc3bfP87/u67qv632/o/d9nd7vXr160bNnT3788UeGDRtW7ntu3rxZIjelTZDr27cvjo6O5W6roljv379PgwYNpJ8bNGhAWlqa2usNGzYkNTW13O0I6rSqsCsL8ypeScc827YxICCAli1b4u3treGoBEFd91ZOVT66fhFu3rwpHQ0Wj221t7cnPj5eutb+LJVKhb6+PjKZrNTl/3TlyhV69uwJgLm5OQqFoszrvVevXqVr164AdO/ene7du0tDYaDoi4OHhwcGBgZkZGSQmZnJoEGDmDdvHn/88QdOTk60adOm1NeeV2RkJE5ORb+XYcOGERERUW5hLz5QqExuqkNpHc61oOt5jaNV53BVlD4SUVedPHmSbt26STe91K1bl3HjxqGvr6/hyAShZngVx7ZWhRjb+mrSqsL+qmnYsCE5OTmkpKRoOhRBqPFelbGtlSXGtr66tOpU/KsgKiqKjh070rRpU6ytrbl8+TImJiaaDksQajxLS0vefvttNmzYwKRJkwgODmbFihW888470s1Za9askc54bdmyhblz57Ju3Tq1sa3PGjp0KOfPn8fT0xOlUsn8+fPL3H/37t05fvw4Y8aMAf4e2wpF1/br1KmDq6sr3bp1w9XVlY8//phx48Yxffr0EmNb//naszZs2MDZs2dJS0tj4sSJdO7cmYCAgBJjWyMjI3nvvffU3vuf//yH9evXM2rUKObMmcPUqVOlsa2ffPKJNFls1apVzJ8/n3379kljW8PCwqr8O3l2bGtQUFCpY1tHjx6Nh4cHMpmMefPmoaenh6enJx999BFjxoyRxrYKladVY1vrNjGlXdPnv95U0506dYoRI0YwdOhQtm/f/tL3rwvjRLWBLuRZjG2tecTYVu32yo5t1dX7v1UqFTKZjL59++Lr64uHh4emQxIEQcuIsa1CMa26xi6T6dZNY48ePcLPz4+1a9cCRXekLly4sMSjJoIgCBURY1uFYlpV2OV6lT8VoQ0KCgo4fvw4R48e1djjJYIgCIJuEedtXrJHjx5x584d2rdvT4MGDTh06BCtWrUSj7AJgiAIL4RWFfYCVeUGNNRU2dnZ9O3bF7lczqlTpzAyMqJt27aaDksQBEHQIVpV2J8+LdB0CP9KnTp1cHZ2platWsjlck2HIwg6LS0tjfDw8HIfUdMVnp6eUh/44ptxQ0NDef3114GiaXRffPEFcrkcpVLJpEmTePvtt4GiS4KrV6/m9OnT1K5dG7lcTkhISJmDWzRh7ty5XLp0iYMHD0qvDRw4kMOHD0t35t++fRt/f38iIiKAomE527ZtQ6FQUFBQwIQJEypsgVuahw8f8uGHH1KnTh3pfqhn3b17l4CAAAoLCzEzM5MeGzx06BBfffUVenp6jB49mlGjRj3np686rSrs2ig6Oprjx49Lz4DOmTNHwxEJwqvBzMzslSjqxRYvXiydAYyNjWXBggV89dVXxMfH8+WXX7J161ZMTU3Jyspi4sSJ1KtXD1tbWzZv3syjR4/49ttvkclkxMXFMXXqVI4ePVoj7rJXKpX8+OOPKBQKEhMTK9Vi9+LFi+zcuZMvv/ySevXqkZ6ejqurK23btqV169ZV2n9oaCjdunXj+vXrpS5fu3YtY8aMYciQIaxcuZL9+/czYsQI1q9fz/79+5HL5Tg7O+Pg4ICpqWmV9v28NP9b02EqlYolS5YQHx+Pl5eXOO0uCC9QREQEP/30ExkZGfz222/MmDGDyMhIEhMTWbBgAc2bN5eO4M6cOcPKlSvR19fHycmJsWPHMnjwYPr160fDhg0ZOXIkwcHBKJVK6ekUS0tLtf2dPXuWNWvWIJfLqVevHqtXr2bGjBmMHTuW7t27k5eXh5OTEz/88ANr167lwoULFBYW4uHhwbBhwwgMDEQul5OZmcnixYuZOXMmOTk55OXlMWfOHDp27MiBAwfYsmULFhYW1K9fHxsbG959913mzJlDcnIyBQUF+Pv7lxgu80+dOnXi1q1bQNHsdn9/f6mo1K1blw8//JDNmzdja2vLnj17OHTokNQnvmvXrnzzzTclivqBAwfYvn07enp6+Pj40L9/f3r27Cl12/P398fd3Z3z58+TnJzM7du3qV+/fqXzU5b//e9/dOjQgTfffJMjR45Uar79jh07mDp1qtQlsGHDhnzzzTclugb6+/uTkZEh/SyXy9m6davaOmFhYVy9erXMwh4bGyt1GbSzs2Pr1q20atWK//u//8PY2FjKaVxc3Et7ckHLCrt2PMl+7949zM3NkclkrF+/nuzsbFHUBZ339U9LSn3dull/3mxaVIhO/bqXe49ulljHzLgFA9oXdWy78dd5LiX/yKjugRXu848//mDXrl18/fXXbNy4kQMHDhAREUFUVBQTJkwAir5gf/zxx+zZswcTExOmTJmCq6srBQUF9OvXj379+hEUFISzszNOTk5ERUWxbt06li5dqravhw8fsnz5ciwtLQkICOD06dM4ODjw448/0r17d86cOUPv3r2Jj48nJSWFnTt3kp+fz8iRI7G3twfAxMSEBQsWcPPmTUaNGoW9vT0xMTFs2rSJNWvWsHLlSiIiIjAyMmLYsGHY2Nhw+PBhzMzMWLRoEQ8ePMDb27vCnuxRUVF06NABKJrw9s9HaIun3D1+/JhatWqVKHj//DkrK4tPP/2UQ4cOkZ+fz6xZs+jfv3+Z+1cqlezatYsDBw5UOj+lTdeDvwfZdOjQgWnTplWqsCclJdG+fftyPxNQ6qn1f6poQl1ubq7UVKZhw4YlptPB31PrXhYtK+w13/z589m8eTP/+9//eO2116RrXIIgvHjW1tbIZDLMzMxo164d+vr6NGrUiKysLGmdBw8eUKtWLekP7caNG6VlHTt2BIoGnMycOROAnj17sn79+hL7atCgAbNnz6awsJDk5GRsbGywt7dny5YtzJo1i+PHj+Pk5ERcXByXLl2Sps49ffpU+qNevL9GjRrx6aefsmXLFvLz8zEyMiIjI4O6devSqFEjAOmoPD4+nosXLxIXFwcUdSPLz88v0fUvKCgIIyMjUlNTad68udSKViaTlRhOo1KppGE2lXnUNikpidatW2NoaIihoSEbNmwgOzu7zPWLP+fAgQMrnZ9/niEByMnJ4cyZM8yfP5+6deuiUCi4evUqVlZWpe63+KxDaZ/5ZSirkevLbvCqXYVdCw7Y27Vrh6WlJTk5OZoORRBeqsocYfdr51LhOm0tetDWokeF6wFqp4vLuh6sp6dX5h/54ptYZTKZ9MdXqVSip6dHfHw8K1euBGD58uUEBwfz+eef06ZNG+nafb169WjcuDFJSUnEx8czf/58fv/9d5ydnZk0aVKZ+/vqq68wNzfnk08+4fLlyyxbtkyt2BbHVPyeyZMnVzhHvfgae3R0NPv27ZMmorVu3ZorV65gYWEhrfvLL7/w+uuvY2xsTEFBAffv35e+UEDRCNoOHTpIMZSXw2JKpbLE56xqfv7pv//9L4WFhbi7uwNFw3SOHDmClZUV9evX5/Hjx9LNc8XT6Yo/c0JCAk2aNJG2lZiYiIWFhVob3Mqciq+IkZEReXl5GBoaqk2ne3ZqXWpqKp07d67Sdv8NrWpQU0vfSNMhlJCVlcWKFSukPr+jR48mOjpadI8ThBqifv36FBYWcu/ePVQqFZMmTSoxsvTZyWw//fQT1tbWdOnSRZpaZm5uTlZWFk2aNOHRo0fExsZKhczBwYHPPvtMaunasWNHoqOjefr0KU+ePGHBggUlYsrIyKBFixZAUfFSKpWYmpqSmZnJw4cPycvL4/z580DR9fLjx48DkJ6eLn3ZKIudnR35+fmcOHECAC8vL8LDw6VJbllZWaxatYqxY8cC4O7uzuLFiykoKHrq6OLFiwQGBqqNvG3dujU3b94kOzubJ0+e4OPjI919n5ubS25ubplT6J4nP8UiIyNZtmwZBw8e5ODBg+zZs4eoqChUKhW2trYcOHAAKDoi3r9/P/369ZM+87p160hPTweKnpD44IMPuHv3rtr2165dqzadrqpFHaBXr14cO3YM+HsSXadOnbh8+TKPHj0iOzubuLg43nrrrSpv+3lp1xF7DbRq1SpWrVqFkZERvr6+yGSyGjsYQxBeVaGhodK12SFDhpR6E1VISAj79u1DLpezaNGiEtsYM2YMbm5utGzZkgkTJhAeHo6dnR329vaEhYVJp++7du1Kz549cXFxQaVSSdPenvXuu+8ya9YsoqKicHd3JzIykoMHD+Lr64u7uzuvvfYa1tbW6OnpMWTIEM6dO4erqyuFhYUlJtCVJigoCD8/P2xtbencuTMzZsxgwoQJ0uNuXl5eUqGZMGECn332GSNHjsTExARjY2M2bNigNnTEyMgIf39/fHx8ABg7diwymQw3NzdGjx5NmzZtyjw9Xtn8nDp1itu3b0s/Z2Rk8Ouvv0rFGqB58+ZYWloSFxeHn58fYWFhuLu7U1hYSI8ePXB1dQWQPvP48eOpXbs2BgYGhISEVPnSaGFhIWPHjuXRo0fcu3cPT09PpkyZgqmpKT/88AP+/v5MmzaNWbNmsXfvXpo2bcqIESOQy+XMnDmT8ePHI5PJ8PPzk26kexm0arpbg2YWtLJopulwUCqV0qmmR48esWnTJvz8/Mq8+UNb6MLUMW2gC3kW092qR1RUFDY2NpiamjJ+/Hj8/Pzo2rWrpsMq1YvOcXZ2Nl988UWlvrjoohc53U2rTsUXqJ5oOgQuXryIra0t0dHRQNE1pJkzZ2p9URcEQfPy8vLw9vbG1dWVFi1a1NiiXh3S0tJwcnLSdBg6QZyKryIDAwNu377N5cuXsbOz03Q4giDokBEjRjBixAhNh6ERLVu21HQIOkO7CrtMM7fFnzt3jpYtW2JhYUGnTp2Ij49Xu9tSEARBEGoKrToVrwnnzp1j6NChfPTRR9JroqgLgiAINZV2HbFrQI8ePfDw8MDNzU3ToQiCIAhChURh/4fc3FwWL16MhYUFU6ZMQU9PjzVr1mg6LEEQBEGoFK06FS+XVf52/+eVk5PD3r172blzp9SwQRAEQVuEh4czePBgPD098fDwwNnZmR9++EFafuXKFby8vBg1ahQjRoxg3bp1am1lDxw4wHvvvYerqyvOzs5ERUVp4mOUKTIyEisrK6nhDkBgYKD0pFKxgQMHSm1vExIS8PT0xMXFhffee49169Y9V5vXp0+fsnz5cmxsbEpdrlQqmTlzJm5ubnh4eJCcnAzA9evXcXV1xdXVldDQ0Crvt6q0qrDL9Kon3Ly8PG7eLBpM0bBhQ77++mu+//77GjGyUBAEoaq8vLzYvn07O3bsYPPmzSxcuJC8vDyysrKYOXMmwcHBfP3113z77bdkZmYSHh4OqI873bNnDxs3bmTFihUkJSVp+BP9LTIyEktLS6nbW0WysrL46KOPmDNnDnv37mXv3r388ssvfP3111Xe9+eff06TJk3K/FIQGRlJvXr12L17N5MnT2bFihUALFy4kODgYPbs2UNWVhYnT56s8r6r4pWvXHl5eQwaNIiCggJOnDhB7dq1pQEGgiDUXBWNbbWxsWHx4sUkJCTw5MkT3NzcGDVqFCkpKQQGBlJYWEjTpk1ZunQpISEh0kjVlStXMnfuXJKTk8nPz8ff358+ffqo7fuvv/6SbqgtKChg6dKlREdH8/jxY6nBiqenJyEhIfz5559s3boVAwMDrK2tCQwMJCIiglOnTpGamsqqVavYunVriTivX79OYGAgxsbGWFtbk5GRwZIlS9i5cyeHDx9GT08Pe3t7xo0bV26eTE1NMTMzIy0tjdOnTzNo0CBp8plMJuPDDz/k7bffZvr06ZUed3rt2jXmzp2LgYEBXbp0YdasWXh6ejJnzhzatm3Ljh07yMjIoEePHmzdupWcnBx69uwJUKn8lCUzM5OEhAQWLVrE5s2bK3Xv0+HDhxk0aJA0YVMul7N06VJq166ttt6GDRs4e/as2muhoaFq3eo8PDyoW7dumVPhYmJipMcVe/XqRXBwMPn5+aSkpEh1xc7OjpiYmHKn4/1bWlXYlYV5L3ybhoaGDBgwAKVS+dIn8AiCLqlpY1u7dOlCs2bNCAoKIi8vD3t7e0aNGiX1SR80aBDLli3jypUrwN8jVQ8cOIBCoWDHjh3cu3cPLy+vEkeHqamp+Pn5YWNjw/79+9m1axfe3t5MmzaNqVOnkpmZSXp6OpaWlgQFBbF3714UCgXTp0/n4sWLANy9e5c9e/aQn59fapzr16/Hz88PBwcHpk+fTu3atUlOTiYqKordu3cD4ObmhqOjI02bNi0zR0lJSaSnp2Nubk5SUhKdOnVSW25kZESjRo1ITU2t9LjTsLAwQkJC6NKlCwEBAaSkpJS5/xs3bnDs2DHS09MrnZ+yOjNGRUUxYMAA+vbty+zZs6UR2eVJSkoqcbBW2ihWX19ffH19y91WRSNcnx3Xqqenh0wm4/79+2o5LB7tWp20qrC/qKfYf/75Z44ePUpQUBBQ9I9UpqFn5AVBeH7ljW2tVasWDx8+xNXVFblcLk3xunbtGiEhIQAEBAQAsHv3brURrsVHl+bm5igUCjIzMzE1NZX2a2ZmRlhYGOHh4Tx69AgrKyuaNGmCTCYjNTWVs2fPYm9vz++//86dO3cYP348AI8fP+bOnTtA0eAZmUxWZpyJiYlS57mBAwcSExPD5cuXuXXrFl5eXkBRG9aUlJQShX3btm0cO3aMrKws8vPzWb58OQqFAplMVuqY1uLJcpUdd3rz5k3pCHjZsmXlrtuuXTsUCkWV8lNWYY+MjGTKlCno6+vj6OjId999J/WvL41MJivzM78MpR0svowDSK0q7C+CSqUiMDCQ8+fP884772BlZSWKuiC8ADVtbOv58+c5d+4c27dvRy6X06VLFwD09fVL/eNaPP8B1P/45ufnk5+fL80PHz9+PMeOHaNPnz64ubkRFRUlTVKzt7fnxIkTnD59mkmTJiGTybC2tmbLli1q+4qIiJD2V1acxdPTQH2E64ABA6SxsWXx8vLCw8OD1NRUvL29adeuHfD3CNd3331XWjc7O5uHDx9iZmZW6XGnehXc7/TsjcfP9j6vbH5K89dff3Hp0iWWLFmCTCYjLy8PY2NjfHx8qF+/fomJfUqlEiMjI1q3bs3ly5fVOvo9ePCA3NxcmjX7e/ZIZU7FV6Rx48akpaXRvn176SywmZkZmZmZ0jrFo12rk1bdPPdvFP/SZTIZa9as4dtvvy1zGpEgCNovIyMDCwsL5HI5x48fp7CwkPz8fKytrTl37hwAa9asKfHH/NkRrnfv3kVPT4/GjRtLoz0HDBggjV1VqVQcP35cbYTryZMnuXXrFlZWVrRq1YrExERpfOjatWu5d+9epeJs0aKFdJng1KlTAFhZWREbG0tubi4qlYqwsDDy8sq+RNm4cWPpzneA4cOHc+LECS5fviyts2rVKpydnYHKjztt06aNtI3g4GASExOpW7eudIo5Li6u1HieJz/FIiMjcXd359ChQxw8eJCoqCgePnzIn3/+ia2tLUeOHJG+UERGRkpH/cWfOSEhASj6ojZv3rwSv3dfX1+1Ea7bt2+v8jS43r17S08RREdH07NnT+RyOa1bt+bChQvA36Ndq9MrccS+du1a1qxZw4kTJ7C0tKRdu3bSN1hBEHRTr1692LRpEx4eHtjb2zNgwADmzZuHv78/QUFB7Nq1iyZNmjB16lQOHTokvW/o0KGcP38eT09PlEplqUfHLi4uLFiwgGbNmkk3jZ0+fZo+ffqQnJws3WxXu3ZtgoODmThxIgqFgg4dOpQ4WisrTl9fX2bPns1XX33F66+/zuPHj2natCleXl64u7ujr6+Pvb19hQOofHx8GD58OO+99x5vvPEGn3/+OfPmzSM7O5uCggL69OnDpEmTgMqPOw0JCWHu3Lno6enRuXNn2rRpg4uLC/Pnz+e1116TZs3/U+vWrSuVn7S0NMLDw9Vyf+TIEZYuXSr9LJPJGDFiBEeOHMHX15fExETc3d1RKBQ0atSIuXPnAlCnTh02bdpEaGgoeXl56OvrM3z4cEaNGlVu3kqzYMECbty4QVZWFp6engwcOBAfHx98fX3ZsGEDTk5OnD17Fjc3NxQKBUuWFN13EhwczNy5c3n69CmdOnWiV69eVd53VWjV2FbTZg1pY9Gyyu/ftWsXixcv5osvvnipw+61jS6ME9UGupBnMba1+v38888YGhrSvn17Nm7ciEqlYvLkyZoOS1LdOV66dCmzZs2qtu3XNC9ybKtWHbHryfQrtZ5SqWTbtm14eXkhl8txc3PjnXfeqfCORkEQhJpCoVAQEhKCoaEhhoaG0jPRr4L8/Hx69+6t6TC0llYVdn39yh0drFixgmXLlpGTk8O0adOQyWSiqAuCoFU6dOjAN998o+kwNEKhUJToHSBUnlYVdlk5D7w9ffpUulPT19eXrKwsvL29X1ZogiAIgvDcnn0K4t/SqrviCwvzS339+vXrODg4SHeOmpiYEBYWVmpjBUEQ/j09PT0xS0EQXqDCwsIKHyOsLK06Yn+qKr1xQk5ODgkJCZw8eZJ+/fq95KgE4dVjYGBAbm4uOTk56Ovr17heEEqlUroZSageIscvhkqlorCwkMLCwhc2n0Srjtifdf36dVJTUwHo2rUrMTExzJkzR8NRCcKrw9jYWOpmVtMkJiZqOgSdJ3L8YshkMhQKBcbGxi9sm9V6xL5o0SIuXbqETCYjODhYrV/v2bNnWblyJfr6+vTr1w8/P79Kb/fnn3/G0dGRwYMHs23bNoAqNxIQBOHfq8kTEGvqo3i6ROS4Zqq2I/bz589z69Yt9u7dy8KFC1m4cKHa8uI+y7t37+bMmTP8/vvvld52x44dGTp0KGPGjHnRYQuCIAiCVqu2wh4TE4O9vT1Q1H7w4cOHZGVlAZCcnIyJiQlNmjRBT0+P/v37ExMTU+E2jxw8XBS0nh5btmzB0dGxusIXBEEQBK1UbefR7t+/r9aLvUGDBqSlpUn9hItH2xUvS05OLnNbxc3xfjx2HB93b7VhDcKL9eTJE02H8EoQea5+IsfVT+S4ehXfnFjVBrEv7QLZv+lcWzxgYXZICL/++uuLCkkoRfHQCaF6iTxXP5Hj6idy/HIolcoKZwI8q9oKe+PGjbl//770c2pqKmZmZqUuq2iMXZ06dWjbti1yubxG3oErCIIgCC+aSqVCqVRWuSd/tRX23r17Ex4ejqurK1evXqVx48ZSW9fmzZuTlZXF7du3sbCwIDo6muXLl5e5LT09vRf6KIAgCIIgaIOqHKkXq9bpbsuXL+fChQvIZDJCQ0O5du0axsbGODg48NNPP0nFfPDgwYwfP766whAEQRCEV4ZWjG0VBEEQBKFytLbznCAIgiAIJYnCLgiCIAg6pEYW9kWLFuHi4oKrqysJCQlqy86ePYuzszMuLi6sX79eQxFqv/JyfO7cOUaPHo2rqytBQUE8fVr68B2hfOXluNiKFSvw9PR8yZHpjvJyfPfuXdzc3HB2dmbu3LkailA3lJfnnTt34uKzcXROAAAJu0lEQVTigpubW4kOo0Ll3bhxA3t7e3bs2FFiWZXrnqqGiY2NVb3//vsqlUql+v3331WjR49WWz5kyBDVnTt3VIWFhSo3NzfVb7/9pokwtVpFOXZwcFDdvXtXpVKpVNOmTVOdOHHipceo7SrKsUqlUv32228qFxcXlYeHx8sOTydUlGN/f3/V999/r1KpVKp58+apUlJSXnqMuqC8PD9+/FhlZ2enUiqVKpVKpfLx8VHFx8drJE5tlp2drfLw8FDNnj1btX379hLLq1r3atwRe3W0ohXUlZdjgIiICCwsLICiroAZGRkaiVObVZRjgCVLljBjxgxNhKcTysvx06dPuXjxIgMHDgQgNDSUpk2baixWbVZenuVyOXK5nJycHAoKCsjNzcXExEST4WolhULBpk2bSu3n8jx1r8YV9vv371O/fn3p5+JWtECprWiLlwmVV16OAanfQGpqKmfOnKF///4vPUZtV1GOIyIi6NGjB82aNdNEeDqhvBw/ePCAOnXqsHjxYtzc3FixYoWmwtR65eW5Vq1a+Pn5YW9vj52dHZ06daJVq1aaClVrGRgYlPm8+vPUvRpX2P9JJZ7Gq3al5Tg9PZ3JkycTGhqq9p9aeD7P5jgzM5OIiAh8fHw0GJHueTbHKpWKe/fu4eXlxY4dO7h27RonTpzQXHA65Nk8Z2VlsXHjRqKiojh+/DiXLl3i+vXrGoxOgBpY2F9kK1qhdOXlGIr+s06cOJEPPviAPn36aCJErVdejs+dO8eDBw9wd3dn6tSpXL16lUWLFmkqVK1VXo7r169P06ZNadGiBfr6+tja2vLbb79pKlStVl6eExMTsbS0pEGDBigUCt566y3RP/4Fe566V+MKe+/evTl27BhAua1oCwoKiI6Opnfv3poMVyuVl2Mouvbr7e1Nv379NBWi1isvx46Ojnz33Xfs27ePdevWYWVlRXBwsCbD1Url5djAwABLS0v++OMPabk4Rfx8ystzs2bNSExMJC8vDygaCtOyZUtNhaqTnqfu1cjOc6IVbfUrK8d9+vShe/fudOnSRVp32LBhuLi4aDBa7VTev+Nit2/fJigoiO3bt2swUu1VXo5v3bpFYGAgKpWKtm3bMm/ePPT0atyxjFYoL8979uwhIiICfX19unTpQkBAgKbD1TpXrlxh6dKlpKSkYGBggLm5OQMHDqR58+bPVfdqZGEXBEEQBOH5iK+vgiAIgqBDRGEXBEEQBB0iCrsgCIIg6BBR2AVBEARBh4jCLgiCIAg6xEDTAQjCq+D27ds4OjqqPUYIEBwczJtvvlnqe8LDwykoKPhX/eRjY2OZMmUKHTp0AODJkyd06NCBkJAQ5HJ5lbZ16tQprl69iq+vL3FxcZiZmWFpacnChQt59913sba2fu44w8PDiYiIoHnz5gAUFBRgYWHB/PnzMTY2LvN99+7dIykpCVtb2+fetyDoGlHYBeEladCggUaeV2/btq20X5VKxYwZM9i7dy8eHh5V2k6/fv2kpkURERE4OTlhaWlJSEjIC4nznXfeUfsS88knn/DZZ5/x0Ucflfme2NhYEhMTRWEXhGeIwi4IGpaYmEhoaCj6+vpkZWXxwQcf0LdvX2l5QUEBs2fP5ubNm8hkMt58801CQ0PJz89n/vz53Lp1i+zsbIYNG8a4cePK3ZdMJqNbt24kJSUBcOLECdavX4+hoSG1a9dmwYIFmJubs3z5cs6dO4dCocDc3JylS5cSGRnJ2bNnefvtt4mKiiIhIYGgoCA+/fRTfH19WbFiBSEhIXTt2hWAsWPH4uPjwxtvvMHHH39Mbm4uOTk5fPjhh/Tq1avCvHTp0oV9+/YBcOHCBZYvX45CoSAvL4/Q0FDq1avH6tWrUalUmJqa4u7uXuV8CIIuEoVdEDTs/v37TJ8+ne7duxMfH8+CBQvUCvuNGze4dOkSR48eBWDfvn08fvyYvXv30rhxY8LCwigsLGT06NH06tWL9u3bl7mvJ0+eEB0djbOzM7m5ucyePZv9+/djYWHBjh07WL16NYGBgezcuZMLFy6gr6/Pd999p9ar2sHBgW3btuHr64utrS2ffvopAMOHD+fYsWN07dqV9PR0EhMT6dOnD76+vowbNw4bGxvS0tJwcXHh+++/x8Cg7D8/BQUFREZG0rlzZ6BocM68efNo3749kZGRbNy4kbVr1zJy5EgKCgrw8fFh8+bNVc6HIOgiUdgF4SV58OABnp6eaq+tWbMGMzMzli1bxqpVq1AqlWRmZqqt06ZNG+rXr8/EiROxs7NjyJAhGBsbExsby19//cVPP/0EQH5+Pn/++WeJQnbjxg21/drZ2eHk5MQvv/xCw4YNsbCwAKBHjx7s2bMHExMT+vbti4eHBw4ODjg5OUnrlGfo0KG4ubkRFBREVFQUjo6O6OvrExsbS3Z2NuvXrweK+rinp6djbm6u9v5Dhw4RFxeHSqXi2rVreHl58f777wPQqFEjli1bxpMnT3j8+HGpM78rmw9B0HWisAvCS1LWNfaZM2cydOhQnJ2duXHjBpMnT1ZbXqtWLXbt2sXVq1elo+3du3ejUCjw8/PD0dGx3P0+e439WTKZTO1nlUolvbZ27VoSExM5efIkHh4ehIeHV/j5im+mS0hI4OjRowQGBgKgUCgIDw9XmyldmmevsU+ePJlmzZpJR/UBAQF8/PHH2NraEh0dzdatW0u8v7L5EARdJx53EwQNu3//Pm+88QYA3333Hfn5+WrLL1++zLfffouVlRVTp07FysqKP/74g27dukmn558+fcrixYtLHO2Xp2XLlqSnp3Pnzh0AYmJi6NSpE8nJyXz55Ze0adOGcePG4eDgUGLGtkwmQ6lUltjm8OHD2b9/Pw8fPpTukn82zgcPHrBw4cIKYwsNDSU8PJy//vpLLUeFhYVERUVJOZLJZBQUFJTYz/PkQxB0hSjsgqBh48aNIyAggPHjx9OtWzdMTExYsmSJtLxFixYcO3YMV1dXvLy8qFevHl27dsXd3R0jIyNcXFwYPXo0xsbGmJqaVnq/hoaGLFy4kBkzZuDp6UlMTAwffPAB5ubmXLt2DWdnZ7y9vUlJSWHw4MFq7+3duzehoaF8//33aq8PHjyYw4cPM3ToUOm1kJAQ/vvf/zJmzBjef/99bGxsKoytSZMmTJw4kTlz5gAwceJEvL29mTx5MiNHjuTu3bt8+eWXvPXWW0RERLB69ep/nQ9B0BViupsgCIIg6BBxxC4IgiAIOkQUdkEQBEHQIaKwC4IgCIIOEYVdEARBEHSIKOyCIAiCoENEYRcEQRAEHSIKuyAIgiDoEFHYBUEQBEGH/D/mjYDVH7LhVgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 576x396 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Results\n",
        "print(classification_report(y_val,rf.predict(X_val)))\n",
        "\n",
        "#plot confusion matrix\n",
        "plot_confusion_matrix(rf, X_val, y_val,xticks_rotation=\"vertical\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "id": "XjcrNQ8SJrbI",
        "outputId": "d24c609a-c4c9-41f2-908b-31d3c9b70351"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.96      0.97      3865\n",
            "           1       0.99      0.99      0.99     12370\n",
            "\n",
            "    accuracy                           0.99     16235\n",
            "   macro avg       0.99      0.98      0.98     16235\n",
            "weighted avg       0.99      0.99      0.99     16235\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning:\n",
            "\n",
            "Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x396 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "varr,fet=[],[]\n",
        "for i in range(len(X_train.columns)):\n",
        "  if rf.feature_importances_[i] > 0.04:\n",
        "    fet.append(rf.feature_importances_[i])\n",
        "    varr.append(X_train.columns[i])\n",
        "import_fet=pd.DataFrame({ \"Varname\": varr, \"Imp\": fet}) \n",
        "import_fet.sort_values(by=\"Imp\", ascending=False)"
      ],
      "metadata": {
        "id": "RPZIvOnjc9q6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "sD1acznhzm8Q"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**XGB Classifier**"
      ],
      "metadata": {
        "id": "mCvzBmbCjG6N"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model_xgb = XGBClassifier(learning_rate=,n_estimators=,max_depth=,\n",
        "                          ,n_jobs=-1)\n",
        "model_xgb.fit(X_train, y_train).score(X_val, y_val)"
      ],
      "metadata": {
        "id": "IOTTdtyAicPf",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:07:12.934437Z",
          "iopub.execute_input": "2021-12-11T01:07:12.934760Z",
          "iopub.status.idle": "2021-12-11T01:08:17.234516Z",
          "shell.execute_reply.started": "2021-12-11T01:07:12.934723Z",
          "shell.execute_reply": "2021-12-11T01:08:17.233820Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "adfd0e94-cee8-4394-a000-b5174dcfa6b1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9844779796735448"
            ]
          },
          "metadata": {},
          "execution_count": 114
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(classification_report(y_val,model_xgb.predict(X_val)))\n",
        "plot_confusion_matrix(rf, X_val, y_val,xticks_rotation=\"vertical\")  \n",
        "plt.show()"
      ],
      "metadata": {
        "id": "u01dnhYyJYvi",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "outputId": "d9e2de95-efa0-45af-b5d9-ef5136991b12"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.95      0.97      3865\n",
            "           1       0.99      0.99      0.99     12370\n",
            "\n",
            "    accuracy                           0.98     16235\n",
            "   macro avg       0.98      0.97      0.98     16235\n",
            "weighted avg       0.98      0.98      0.98     16235\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning:\n",
            "\n",
            "Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x396 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Decision Tree Classifier**"
      ],
      "metadata": {
        "id": "2qmga3ehjLKY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "dt = DecisionTreeClassifier(max_depth=12,min_samples_leaf=3,min_samples_split=7,random_state=42)\n",
        "dt.fit(X_train, y_train).score(X_val, y_val)"
      ],
      "metadata": {
        "id": "1KboZlB7hGTU",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:08:17.235904Z",
          "iopub.execute_input": "2021-12-11T01:08:17.236295Z",
          "iopub.status.idle": "2021-12-11T01:08:17.925387Z",
          "shell.execute_reply.started": "2021-12-11T01:08:17.236257Z",
          "shell.execute_reply": "2021-12-11T01:08:17.924525Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d6d856e6-b7d4-4207-f87d-4e30eedf3066"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.983615645210964"
            ]
          },
          "metadata": {},
          "execution_count": 119
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(classification_report(y_val,dt.predict(X_val)))\n",
        "plot_confusion_matrix(rf, X_val, y_val,xticks_rotation=\"vertical\")  \n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "id": "0oT-CFxhJlHa",
        "outputId": "dd78da34-089f-4fdf-bea6-8d76f1ec856a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      0.96      0.97      3865\n",
            "           1       0.99      0.99      0.99     12370\n",
            "\n",
            "    accuracy                           0.98     16235\n",
            "   macro avg       0.98      0.98      0.98     16235\n",
            "weighted avg       0.98      0.98      0.98     16235\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning:\n",
            "\n",
            "Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x396 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**SVM**"
      ],
      "metadata": {
        "id": "stE0YFT0jiv9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "svm_cls= svm.SVC()\n",
        "svm_cls.fit(X_train,y_train)\n",
        "predicted_svm = svm_cls.predict(X_val)\n",
        "print(\"Accuracy for SVM :-> \",np.mean(predicted_svm == y_val))\n",
        "print(metrics.classification_report(y_val, predicted_svm))"
      ],
      "metadata": {
        "id": "kWzRZoD0jila",
        "execution": {
          "iopub.status.busy": "2021-12-11T01:17:03.046813Z",
          "iopub.execute_input": "2021-12-11T01:17:03.047187Z",
          "iopub.status.idle": "2021-12-11T01:24:28.865956Z",
          "shell.execute_reply.started": "2021-12-11T01:17:03.047148Z",
          "shell.execute_reply": "2021-12-11T01:24:28.864748Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "914bae63-75fc-444e-9bdb-530281b8dd3a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy for SVM :->  0.9764705882352941\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.90      0.95      3865\n",
            "           1       0.97      1.00      0.98     12370\n",
            "\n",
            "    accuracy                           0.98     16235\n",
            "   macro avg       0.98      0.95      0.97     16235\n",
            "weighted avg       0.98      0.98      0.98     16235\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "H9xwCQ2lJ6Iy"
      },
      "source": [
        "**Bagging**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bdL-jf6P8w5w",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ff366e71-f468-43bb-aa4d-9e85ef9f62f5"
      },
      "source": [
        "bag = BaggingClassifier(svm_cls,random_state=0)\n",
        "bag.fit(X_train, y_train).score(X_val,y_val)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9764705882352941"
            ]
          },
          "metadata": {},
          "execution_count": 124
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(classification_report(y_val,bag.predict(X_val)))\n",
        "plot_confusion_matrix(rf, X_val, y_val,xticks_rotation=\"vertical\")  \n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "id": "oPqsVefHJwMh",
        "outputId": "43c2b7f9-77af-4ac5-aa50-50e7e9d47c06"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.90      0.95      3865\n",
            "           1       0.97      1.00      0.98     12370\n",
            "\n",
            "    accuracy                           0.98     16235\n",
            "   macro avg       0.98      0.95      0.97     16235\n",
            "weighted avg       0.98      0.98      0.98     16235\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning:\n",
            "\n",
            "Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x396 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zDzPyOI2Xwgt"
      },
      "source": [
        "**Stacking**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LfPoIPsRN8Fm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3201c5e9-193b-4299-8b0c-82d918cd95c2"
      },
      "source": [
        "models = [\n",
        "    (\"bag\", BaggingClassifier(svm_cls, random_state=0)),\n",
        "    (\"xgb\",XGBClassifier()),\n",
        "    (\"dt\", make_pipeline(DecisionTreeClassifier(max_depth=15))),\n",
        "    ('rf', RandomForestClassifier())\n",
        "    ]\n",
        "\n",
        "stacking = StackingClassifier(estimators=models,final_estimator=RandomForestClassifier(),cv=5)\n",
        "stacking.fit(X_train, y_train).score( X_val, y_val)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9833692639359408"
            ]
          },
          "metadata": {},
          "execution_count": 130
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(classification_report(y_val,stacking.predict(X_val)))\n",
        "plot_confusion_matrix(rf, X_val, y_val,xticks_rotation=\"vertical\")  \n",
        "plt.show()"
      ],
      "metadata": {
        "id": "TWTLjt7zJ2Fp",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "outputId": "0b4d9007-929e-4aaf-d2af-64490316ba4d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      0.96      0.97      3865\n",
            "           1       0.99      0.99      0.99     12370\n",
            "\n",
            "    accuracy                           0.98     16235\n",
            "   macro avg       0.98      0.98      0.98     16235\n",
            "weighted avg       0.98      0.98      0.98     16235\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning:\n",
            "\n",
            "Function plot_confusion_matrix is deprecated; Function `plot_confusion_matrix` is deprecated in 1.0 and will be removed in 1.2. Use one of the class methods: ConfusionMatrixDisplay.from_predictions or ConfusionMatrixDisplay.from_estimator.\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x396 with 2 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Multiclass Model**"
      ],
      "metadata": {
        "id": "TqQ6hGQcX-IP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "uIGclJz7YDkf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "zKBNKlDFYETd"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "tG1JXO8QYEs0"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}