Credit_card_fraud_detection.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Credit card fraud detection.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C6M6VTfQml2z"
      },
      "source": [
        "## Importing the libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iuaT8l0kmZvw"
      },
      "source": [
        "import numpy as np\r\n",
        "import pandas as pd\r\n",
        "import matplotlib.pyplot as plt\r\n",
        "import seaborn as sns\r\n",
        "from matplotlib import gridspec"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LJSv5zRTnM4g"
      },
      "source": [
        "## Importing Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EnQoJn3VmzbP"
      },
      "source": [
        "dataset = pd.read_csv(\"/content/drive/MyDrive/creditcard.csv\")"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rg4F33xAnS0Q"
      },
      "source": [
        "## Understanding the Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 220
        },
        "id": "RQ3KaDVfnKXy",
        "outputId": "466b0476-4e63-4239-8f3e-df6cc33ee00a"
      },
      "source": [
        "dataset.head()"
      ],
      "execution_count": 11,
      "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>Time</th>\n",
              "      <th>V1</th>\n",
              "      <th>V2</th>\n",
              "      <th>V3</th>\n",
              "      <th>V4</th>\n",
              "      <th>V5</th>\n",
              "      <th>V6</th>\n",
              "      <th>V7</th>\n",
              "      <th>V8</th>\n",
              "      <th>V9</th>\n",
              "      <th>V10</th>\n",
              "      <th>V11</th>\n",
              "      <th>V12</th>\n",
              "      <th>V13</th>\n",
              "      <th>V14</th>\n",
              "      <th>V15</th>\n",
              "      <th>V16</th>\n",
              "      <th>V17</th>\n",
              "      <th>V18</th>\n",
              "      <th>V19</th>\n",
              "      <th>V20</th>\n",
              "      <th>V21</th>\n",
              "      <th>V22</th>\n",
              "      <th>V23</th>\n",
              "      <th>V24</th>\n",
              "      <th>V25</th>\n",
              "      <th>V26</th>\n",
              "      <th>V27</th>\n",
              "      <th>V28</th>\n",
              "      <th>Amount</th>\n",
              "      <th>Class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.0</td>\n",
              "      <td>-1.359807</td>\n",
              "      <td>-0.072781</td>\n",
              "      <td>2.536347</td>\n",
              "      <td>1.378155</td>\n",
              "      <td>-0.338321</td>\n",
              "      <td>0.462388</td>\n",
              "      <td>0.239599</td>\n",
              "      <td>0.098698</td>\n",
              "      <td>0.363787</td>\n",
              "      <td>0.090794</td>\n",
              "      <td>-0.551600</td>\n",
              "      <td>-0.617801</td>\n",
              "      <td>-0.991390</td>\n",
              "      <td>-0.311169</td>\n",
              "      <td>1.468177</td>\n",
              "      <td>-0.470401</td>\n",
              "      <td>0.207971</td>\n",
              "      <td>0.025791</td>\n",
              "      <td>0.403993</td>\n",
              "      <td>0.251412</td>\n",
              "      <td>-0.018307</td>\n",
              "      <td>0.277838</td>\n",
              "      <td>-0.110474</td>\n",
              "      <td>0.066928</td>\n",
              "      <td>0.128539</td>\n",
              "      <td>-0.189115</td>\n",
              "      <td>0.133558</td>\n",
              "      <td>-0.021053</td>\n",
              "      <td>149.62</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.191857</td>\n",
              "      <td>0.266151</td>\n",
              "      <td>0.166480</td>\n",
              "      <td>0.448154</td>\n",
              "      <td>0.060018</td>\n",
              "      <td>-0.082361</td>\n",
              "      <td>-0.078803</td>\n",
              "      <td>0.085102</td>\n",
              "      <td>-0.255425</td>\n",
              "      <td>-0.166974</td>\n",
              "      <td>1.612727</td>\n",
              "      <td>1.065235</td>\n",
              "      <td>0.489095</td>\n",
              "      <td>-0.143772</td>\n",
              "      <td>0.635558</td>\n",
              "      <td>0.463917</td>\n",
              "      <td>-0.114805</td>\n",
              "      <td>-0.183361</td>\n",
              "      <td>-0.145783</td>\n",
              "      <td>-0.069083</td>\n",
              "      <td>-0.225775</td>\n",
              "      <td>-0.638672</td>\n",
              "      <td>0.101288</td>\n",
              "      <td>-0.339846</td>\n",
              "      <td>0.167170</td>\n",
              "      <td>0.125895</td>\n",
              "      <td>-0.008983</td>\n",
              "      <td>0.014724</td>\n",
              "      <td>2.69</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.0</td>\n",
              "      <td>-1.358354</td>\n",
              "      <td>-1.340163</td>\n",
              "      <td>1.773209</td>\n",
              "      <td>0.379780</td>\n",
              "      <td>-0.503198</td>\n",
              "      <td>1.800499</td>\n",
              "      <td>0.791461</td>\n",
              "      <td>0.247676</td>\n",
              "      <td>-1.514654</td>\n",
              "      <td>0.207643</td>\n",
              "      <td>0.624501</td>\n",
              "      <td>0.066084</td>\n",
              "      <td>0.717293</td>\n",
              "      <td>-0.165946</td>\n",
              "      <td>2.345865</td>\n",
              "      <td>-2.890083</td>\n",
              "      <td>1.109969</td>\n",
              "      <td>-0.121359</td>\n",
              "      <td>-2.261857</td>\n",
              "      <td>0.524980</td>\n",
              "      <td>0.247998</td>\n",
              "      <td>0.771679</td>\n",
              "      <td>0.909412</td>\n",
              "      <td>-0.689281</td>\n",
              "      <td>-0.327642</td>\n",
              "      <td>-0.139097</td>\n",
              "      <td>-0.055353</td>\n",
              "      <td>-0.059752</td>\n",
              "      <td>378.66</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1.0</td>\n",
              "      <td>-0.966272</td>\n",
              "      <td>-0.185226</td>\n",
              "      <td>1.792993</td>\n",
              "      <td>-0.863291</td>\n",
              "      <td>-0.010309</td>\n",
              "      <td>1.247203</td>\n",
              "      <td>0.237609</td>\n",
              "      <td>0.377436</td>\n",
              "      <td>-1.387024</td>\n",
              "      <td>-0.054952</td>\n",
              "      <td>-0.226487</td>\n",
              "      <td>0.178228</td>\n",
              "      <td>0.507757</td>\n",
              "      <td>-0.287924</td>\n",
              "      <td>-0.631418</td>\n",
              "      <td>-1.059647</td>\n",
              "      <td>-0.684093</td>\n",
              "      <td>1.965775</td>\n",
              "      <td>-1.232622</td>\n",
              "      <td>-0.208038</td>\n",
              "      <td>-0.108300</td>\n",
              "      <td>0.005274</td>\n",
              "      <td>-0.190321</td>\n",
              "      <td>-1.175575</td>\n",
              "      <td>0.647376</td>\n",
              "      <td>-0.221929</td>\n",
              "      <td>0.062723</td>\n",
              "      <td>0.061458</td>\n",
              "      <td>123.50</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2.0</td>\n",
              "      <td>-1.158233</td>\n",
              "      <td>0.877737</td>\n",
              "      <td>1.548718</td>\n",
              "      <td>0.403034</td>\n",
              "      <td>-0.407193</td>\n",
              "      <td>0.095921</td>\n",
              "      <td>0.592941</td>\n",
              "      <td>-0.270533</td>\n",
              "      <td>0.817739</td>\n",
              "      <td>0.753074</td>\n",
              "      <td>-0.822843</td>\n",
              "      <td>0.538196</td>\n",
              "      <td>1.345852</td>\n",
              "      <td>-1.119670</td>\n",
              "      <td>0.175121</td>\n",
              "      <td>-0.451449</td>\n",
              "      <td>-0.237033</td>\n",
              "      <td>-0.038195</td>\n",
              "      <td>0.803487</td>\n",
              "      <td>0.408542</td>\n",
              "      <td>-0.009431</td>\n",
              "      <td>0.798278</td>\n",
              "      <td>-0.137458</td>\n",
              "      <td>0.141267</td>\n",
              "      <td>-0.206010</td>\n",
              "      <td>0.502292</td>\n",
              "      <td>0.219422</td>\n",
              "      <td>0.215153</td>\n",
              "      <td>69.99</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Time        V1        V2        V3  ...       V27       V28  Amount  Class\n",
              "0   0.0 -1.359807 -0.072781  2.536347  ...  0.133558 -0.021053  149.62      0\n",
              "1   0.0  1.191857  0.266151  0.166480  ... -0.008983  0.014724    2.69      0\n",
              "2   1.0 -1.358354 -1.340163  1.773209  ... -0.055353 -0.059752  378.66      0\n",
              "3   1.0 -0.966272 -0.185226  1.792993  ...  0.062723  0.061458  123.50      0\n",
              "4   2.0 -1.158233  0.877737  1.548718  ...  0.219422  0.215153   69.99      0\n",
              "\n",
              "[5 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AOtjLvqunaK3",
        "outputId": "699392ea-5451-40e3-c271-179c3c9defe2"
      },
      "source": [
        "dataset.shape"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(284807, 31)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ptmy7WXxneF2",
        "outputId": "2674e459-d068-44e1-e9ad-9b1216034d32"
      },
      "source": [
        "print(dataset.describe())"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "                Time            V1  ...         Amount          Class\n",
            "count  284807.000000  2.848070e+05  ...  284807.000000  284807.000000\n",
            "mean    94813.859575  3.919560e-15  ...      88.349619       0.001727\n",
            "std     47488.145955  1.958696e+00  ...     250.120109       0.041527\n",
            "min         0.000000 -5.640751e+01  ...       0.000000       0.000000\n",
            "25%     54201.500000 -9.203734e-01  ...       5.600000       0.000000\n",
            "50%     84692.000000  1.810880e-02  ...      22.000000       0.000000\n",
            "75%    139320.500000  1.315642e+00  ...      77.165000       0.000000\n",
            "max    172792.000000  2.454930e+00  ...   25691.160000       1.000000\n",
            "\n",
            "[8 rows x 31 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WCjpD6yQpMpR"
      },
      "source": [
        "# This dataset is highly unbalanced so for this first of all we will try building our model.\r\n",
        "# then if we get bad results, we will use some different technique to solve this issue"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KQMloAHjq3Rv"
      },
      "source": [
        "# let's print the amount details for fruadulent and actual transactions and then comparing the mean (average) in both cases"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z3v1BTne-I9d"
      },
      "source": [
        "fraud = dataset[dataset['Class']==1]\r\n",
        "\r\n",
        "valid = dataset[dataset['Class']==0]"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CM7bK2jMracW",
        "outputId": "73efb555-ff7d-476e-b1da-394dd9d2276e"
      },
      "source": [
        "print(\"Amount transactions in case of fradulent:\")\r\n",
        "fraud.Amount.describe()"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Amount transactions in case of fradulent:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count     492.000000\n",
              "mean      122.211321\n",
              "std       256.683288\n",
              "min         0.000000\n",
              "25%         1.000000\n",
              "50%         9.250000\n",
              "75%       105.890000\n",
              "max      2125.870000\n",
              "Name: Amount, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5FiQISrerv5m",
        "outputId": "d593910f-86b4-44cf-fe21-a920995b5d3e"
      },
      "source": [
        "print(\"Amount transactions in case of actual:\")\r\n",
        "valid.Amount.describe()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Amount transactions in case of actual:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    284315.000000\n",
              "mean         88.291022\n",
              "std         250.105092\n",
              "min           0.000000\n",
              "25%           5.650000\n",
              "50%          22.000000\n",
              "75%          77.050000\n",
              "max       25691.160000\n",
              "Name: Amount, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LseiskCqr479"
      },
      "source": [
        "# Therefore, we can clearly see that our fraudulent average transactions are more than the actual transactions\r\n",
        " "
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gD1z49mSsRV6"
      },
      "source": [
        "## Visualizing the Correlation matrix"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 570
        },
        "id": "L5ps0RO7sHpe",
        "outputId": "a5cb44cc-e385-4f6e-fd10-19b79d787705"
      },
      "source": [
        "# We use correlation matrix to get best features we can work with in our future data modelling\r\n",
        "\r\n",
        "corr = dataset.corr()\r\n",
        "fig = plt.figure(figsize=(12,9))\r\n",
        "sns.heatmap(corr  , vmax = 0.8 , square = True)\r\n",
        "plt.show()"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x648 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "188CvwwWYKA_"
      },
      "source": [
        ""
      ],
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QUAfix_7uAYx"
      },
      "source": [
        "## Splitting the dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lxaPNiVOYbhl",
        "outputId": "f62bda45-73a2-42e1-c737-ec39de74c6a7"
      },
      "source": [
        "# we will take small amount of dataset\r\n",
        "\r\n",
        "data1 = dataset.sample(frac = 0.1, random_state=1)\r\n",
        "data1.shape"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(28481, 31)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NirpHHiLYyjM",
        "outputId": "3d19b88e-886b-4e59-84f9-240a4e097601"
      },
      "source": [
        "dataset.shape"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(284807, 31)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TixWLxbonm43",
        "outputId": "3b1375df-0424-414b-feeb-cdd7f20ec45e"
      },
      "source": [
        "#lets check number of fraud and valid cases in our currrent dataset\r\n",
        "\r\n",
        "fraud = data1[data1['Class'] == 1] \r\n",
        "valid = data1[data1['Class'] == 0] \r\n",
        "\r\n",
        "outlier_fraction = len(fraud)/float(len(valid))\r\n",
        "\r\n",
        "print(outlier_fraction)\r\n",
        "\r\n",
        "print(\"Fraud cases: {}\".format(len(fraud)))\r\n",
        "print(\"Valid cases: {}\".format(len(valid)))"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.0017234102419808666\n",
            "Fraud cases: 49\n",
            "Valid cases: 28432\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 220
        },
        "id": "T6HE-w0XY1D1",
        "outputId": "ff2a039f-5f0e-4054-fbad-0c9b88ff8883"
      },
      "source": [
        "data1.head()"
      ],
      "execution_count": 24,
      "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>Time</th>\n",
              "      <th>V1</th>\n",
              "      <th>V2</th>\n",
              "      <th>V3</th>\n",
              "      <th>V4</th>\n",
              "      <th>V5</th>\n",
              "      <th>V6</th>\n",
              "      <th>V7</th>\n",
              "      <th>V8</th>\n",
              "      <th>V9</th>\n",
              "      <th>V10</th>\n",
              "      <th>V11</th>\n",
              "      <th>V12</th>\n",
              "      <th>V13</th>\n",
              "      <th>V14</th>\n",
              "      <th>V15</th>\n",
              "      <th>V16</th>\n",
              "      <th>V17</th>\n",
              "      <th>V18</th>\n",
              "      <th>V19</th>\n",
              "      <th>V20</th>\n",
              "      <th>V21</th>\n",
              "      <th>V22</th>\n",
              "      <th>V23</th>\n",
              "      <th>V24</th>\n",
              "      <th>V25</th>\n",
              "      <th>V26</th>\n",
              "      <th>V27</th>\n",
              "      <th>V28</th>\n",
              "      <th>Amount</th>\n",
              "      <th>Class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>169876</th>\n",
              "      <td>119907.0</td>\n",
              "      <td>-0.611712</td>\n",
              "      <td>-0.769705</td>\n",
              "      <td>-0.149759</td>\n",
              "      <td>-0.224877</td>\n",
              "      <td>2.028577</td>\n",
              "      <td>-2.019887</td>\n",
              "      <td>0.292491</td>\n",
              "      <td>-0.523020</td>\n",
              "      <td>0.358468</td>\n",
              "      <td>0.070050</td>\n",
              "      <td>-0.854023</td>\n",
              "      <td>0.547347</td>\n",
              "      <td>0.616448</td>\n",
              "      <td>-0.101785</td>\n",
              "      <td>-0.608492</td>\n",
              "      <td>-0.288559</td>\n",
              "      <td>-0.606199</td>\n",
              "      <td>-0.900746</td>\n",
              "      <td>-0.201311</td>\n",
              "      <td>-0.196039</td>\n",
              "      <td>-0.075208</td>\n",
              "      <td>0.045536</td>\n",
              "      <td>0.380739</td>\n",
              "      <td>0.023440</td>\n",
              "      <td>-2.220686</td>\n",
              "      <td>-0.201146</td>\n",
              "      <td>0.066501</td>\n",
              "      <td>0.221180</td>\n",
              "      <td>1.79</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>127467</th>\n",
              "      <td>78340.0</td>\n",
              "      <td>-0.814682</td>\n",
              "      <td>1.319219</td>\n",
              "      <td>1.329415</td>\n",
              "      <td>0.027273</td>\n",
              "      <td>-0.284871</td>\n",
              "      <td>-0.653985</td>\n",
              "      <td>0.321552</td>\n",
              "      <td>0.435975</td>\n",
              "      <td>-0.704298</td>\n",
              "      <td>-0.600684</td>\n",
              "      <td>0.096501</td>\n",
              "      <td>0.710309</td>\n",
              "      <td>0.778924</td>\n",
              "      <td>0.353867</td>\n",
              "      <td>0.953048</td>\n",
              "      <td>-0.103542</td>\n",
              "      <td>0.128899</td>\n",
              "      <td>-0.964070</td>\n",
              "      <td>-0.477130</td>\n",
              "      <td>-0.009041</td>\n",
              "      <td>-0.128619</td>\n",
              "      <td>-0.368565</td>\n",
              "      <td>0.090660</td>\n",
              "      <td>0.401147</td>\n",
              "      <td>-0.261034</td>\n",
              "      <td>0.080621</td>\n",
              "      <td>0.162427</td>\n",
              "      <td>0.059456</td>\n",
              "      <td>1.98</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>137900</th>\n",
              "      <td>82382.0</td>\n",
              "      <td>-0.318193</td>\n",
              "      <td>1.118618</td>\n",
              "      <td>0.969864</td>\n",
              "      <td>-0.127052</td>\n",
              "      <td>0.569563</td>\n",
              "      <td>-0.532484</td>\n",
              "      <td>0.706252</td>\n",
              "      <td>-0.064966</td>\n",
              "      <td>-0.463271</td>\n",
              "      <td>-0.528357</td>\n",
              "      <td>-0.813214</td>\n",
              "      <td>0.111898</td>\n",
              "      <td>1.133875</td>\n",
              "      <td>-0.649766</td>\n",
              "      <td>0.861035</td>\n",
              "      <td>0.583787</td>\n",
              "      <td>-0.344094</td>\n",
              "      <td>0.028312</td>\n",
              "      <td>0.219760</td>\n",
              "      <td>0.206028</td>\n",
              "      <td>-0.305402</td>\n",
              "      <td>-0.774704</td>\n",
              "      <td>-0.123884</td>\n",
              "      <td>-0.495687</td>\n",
              "      <td>-0.018148</td>\n",
              "      <td>0.121679</td>\n",
              "      <td>0.249050</td>\n",
              "      <td>0.092516</td>\n",
              "      <td>0.89</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21513</th>\n",
              "      <td>31717.0</td>\n",
              "      <td>-1.328271</td>\n",
              "      <td>1.018378</td>\n",
              "      <td>1.775426</td>\n",
              "      <td>-1.574193</td>\n",
              "      <td>-0.117696</td>\n",
              "      <td>-0.457733</td>\n",
              "      <td>0.681867</td>\n",
              "      <td>-0.031641</td>\n",
              "      <td>0.383872</td>\n",
              "      <td>0.334853</td>\n",
              "      <td>0.905210</td>\n",
              "      <td>-0.277872</td>\n",
              "      <td>-1.336179</td>\n",
              "      <td>-0.012238</td>\n",
              "      <td>0.212828</td>\n",
              "      <td>0.769345</td>\n",
              "      <td>-1.010534</td>\n",
              "      <td>0.357288</td>\n",
              "      <td>-0.125585</td>\n",
              "      <td>0.315823</td>\n",
              "      <td>-0.220815</td>\n",
              "      <td>-0.419013</td>\n",
              "      <td>-0.239197</td>\n",
              "      <td>0.009967</td>\n",
              "      <td>0.232829</td>\n",
              "      <td>0.814177</td>\n",
              "      <td>0.098797</td>\n",
              "      <td>-0.004273</td>\n",
              "      <td>15.98</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>134700</th>\n",
              "      <td>80923.0</td>\n",
              "      <td>1.276712</td>\n",
              "      <td>0.617120</td>\n",
              "      <td>-0.578014</td>\n",
              "      <td>0.879173</td>\n",
              "      <td>0.061706</td>\n",
              "      <td>-1.472002</td>\n",
              "      <td>0.373692</td>\n",
              "      <td>-0.287204</td>\n",
              "      <td>-0.084482</td>\n",
              "      <td>-0.696578</td>\n",
              "      <td>0.101512</td>\n",
              "      <td>-0.433200</td>\n",
              "      <td>-0.634867</td>\n",
              "      <td>-1.188354</td>\n",
              "      <td>1.028063</td>\n",
              "      <td>0.576121</td>\n",
              "      <td>1.068353</td>\n",
              "      <td>0.332690</td>\n",
              "      <td>-0.362811</td>\n",
              "      <td>-0.105286</td>\n",
              "      <td>-0.160161</td>\n",
              "      <td>-0.430404</td>\n",
              "      <td>-0.076738</td>\n",
              "      <td>0.258708</td>\n",
              "      <td>0.552170</td>\n",
              "      <td>0.370701</td>\n",
              "      <td>-0.034255</td>\n",
              "      <td>0.041709</td>\n",
              "      <td>0.76</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            Time        V1        V2  ...       V28  Amount  Class\n",
              "169876  119907.0 -0.611712 -0.769705  ...  0.221180    1.79      0\n",
              "127467   78340.0 -0.814682  1.319219  ...  0.059456    1.98      0\n",
              "137900   82382.0 -0.318193  1.118618  ...  0.092516    0.89      0\n",
              "21513    31717.0 -1.328271  1.018378  ... -0.004273   15.98      0\n",
              "134700   80923.0  1.276712  0.617120  ...  0.041709    0.76      0\n",
              "\n",
              "[5 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XyyRZA2TZpW5",
        "outputId": "1f3a7c87-bcac-4cf5-a566-085c85e7ab4b"
      },
      "source": [
        "# Splitting the sample dataset\r\n",
        "\r\n",
        "x = data1.iloc[:,:-1]\r\n",
        "y = data1.iloc[:,-1]\r\n",
        "\r\n",
        "print(x.shape)\r\n",
        "print(y.shape)"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(28481, 30)\n",
            "(28481,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-6dyFRzJaJqe"
      },
      "source": [
        "## Training Our Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pKX5W6c5aEoo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7d0de0e1-e4ad-4268-b137-55da35bf0d3e"
      },
      "source": [
        "# We will be using two different algorithms:-\r\n",
        "# 1) Isolation Forest Algorithm\r\n",
        "# 2) Local Outlier Factor(LOF) Algorithm\r\n",
        "\r\n",
        "from sklearn.ensemble import IsolationForest\r\n",
        "from sklearn.neighbors import LocalOutlierFactor\r\n",
        "from sklearn.metrics import classification_report,accuracy_score,confusion_matrix\r\n",
        "\r\n",
        "state = np.random.RandomState(42)\r\n",
        "\r\n",
        "clf_1 = IsolationForest(n_estimators=100, max_samples=len(x), \r\n",
        "                                       contamination=outlier_fraction,random_state=state, verbose=0)\r\n",
        "\r\n",
        "clf_1.fit(x)\r\n",
        "y_pred_1 = clf_1.predict(x)\r\n",
        "\r\n",
        "y_pred_1[y_pred_1 == 1] = 0\r\n",
        "y_pred_1[y_pred_1 == -1] = 1\r\n",
        "\r\n",
        "\r\n",
        "n_errors = (y_pred_1 != y).sum()\r\n",
        "\r\n",
        "print(\"Isolation Forest:{}\".format(n_errors))\r\n",
        "print(\"Accuracy score: {}\".format(accuracy_score(y,y_pred_1)))\r\n",
        "print(\"{}\".format(classification_report(y,y_pred_1)))\r\n",
        "\r\n"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Isolation Forest:77\n",
            "Accuracy score: 0.9972964432428637\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      1.00      1.00     28432\n",
            "           1       0.22      0.22      0.22        49\n",
            "\n",
            "    accuracy                           1.00     28481\n",
            "   macro avg       0.61      0.61      0.61     28481\n",
            "weighted avg       1.00      1.00      1.00     28481\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EjDTvfz0-VUo",
        "outputId": "0d1e701d-2fbb-44d2-dce7-8367ee0e5682"
      },
      "source": [
        "clf_2 = LocalOutlierFactor(n_neighbors=20, algorithm='auto', \r\n",
        "                                              leaf_size=30, metric='minkowski',\r\n",
        "                                              p=2, metric_params=None, contamination=outlier_fraction)\r\n",
        "\r\n",
        "y_pred_2 = clf_2.fit_predict(x)\r\n",
        "\r\n",
        "scores_prediction = clf_2.negative_outlier_factor_\r\n",
        "print(scores_prediction)\r\n",
        "\r\n",
        "y_pred_2[y_pred_2 == 1] = 0\r\n",
        "y_pred_2[y_pred_2 == -1] = 1\r\n",
        "\r\n",
        "\r\n",
        "n_errors = (y_pred_2 != y).sum()\r\n",
        "\r\n",
        "print(\"Isolation Forest:{}\".format(n_errors))\r\n",
        "print(\"Accuracy score: {}\".format(accuracy_score(y,y_pred_2)))\r\n",
        "print(\"{}\".format(classification_report(y,y_pred_2)))"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[-1.04009684 -1.01681044 -1.02971646 ... -0.9980937  -1.05205114\n",
            " -0.99057752]\n",
            "Isolation Forest:97\n",
            "Accuracy score: 0.9965942207085425\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      1.00      1.00     28432\n",
            "           1       0.02      0.02      0.02        49\n",
            "\n",
            "    accuracy                           1.00     28481\n",
            "   macro avg       0.51      0.51      0.51     28481\n",
            "weighted avg       1.00      1.00      1.00     28481\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K_5KubMQBnii"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}