DAI_ASSIGNMENT_1_Q4_data_method_Mitigation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "DAI_ASSIGNMENT_1_Q4_data_method_Mitigation.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "5b95b5621457464daaab970dc0158d5a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_b221d97ce2004f1388ea7a507232015b",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_28032979b44043e3b6a3e3668791da3c",
              "IPY_MODEL_d715ccd4fa2349079418a9498c943638"
            ]
          }
        },
        "b221d97ce2004f1388ea7a507232015b": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "28032979b44043e3b6a3e3668791da3c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_15fb97ef90934fd4bbcb271315f44b9c",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 18983,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 18983,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_a34a6b326d3e46e9b02cf2afd7387e4e"
          }
        },
        "d715ccd4fa2349079418a9498c943638": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_6ec468aad5914fba9d7acc0cc0c696c2",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 18983/18983 [05:58&lt;00:00, 52.98it/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_7e49534410ce4de09893103761712024"
          }
        },
        "15fb97ef90934fd4bbcb271315f44b9c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "a34a6b326d3e46e9b02cf2afd7387e4e": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "6ec468aad5914fba9d7acc0cc0c696c2": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "7e49534410ce4de09893103761712024": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rRAbTuv6sVQA"
      },
      "source": [
        "Q4. ​Dataset​: Labeled Faces in the Wild, [​Link​ for dataset] ​[50 Marks + 10 Marks Bonus]\n",
        "You need to perform binary classification for a person wearing ‘sunglasses’/’Eyeglasses’ in the dataset. There are number of attributes present for LFW dataset(please refer the link for details about the attributes labeling).[​Link​ for attributes]\n",
        "● Separate the dataset into 2 files. 65% and 35% of the total dataset.\n",
        "● For 65% dataset, perform binary classification for the person wearing ‘sunglasses’ /’Eyeglasses’\n",
        "or not. [​Note​: In the dataset, eyeglasses and sunglasses are treated as different entities, you need\n",
        "to consider them as single entity]\n",
        "● UseDenseLayerModel:[128--128--128--64--1]\n",
        " \n",
        "● Keep the loss function as the “mean square error”.\n",
        "● Do you think there is any kind of bias in the system? Evaluate the system using 3 different\n",
        "evaluation metrics to see if there is any bias or not. [10 Marks]\n",
        "● Come up with a new evaluation metric to detect if there is a bias in the system.\n",
        "[ Bonus 10 Marks]\n",
        "● If you observed any bias in the system, mitigate the bias by: [30 Marks]\n",
        "○ DATA method (Training using more data): You may use more data for training from that\n",
        "35% data. Report the accuracy after mitigation and compare it with previous\n",
        "classification results.\n",
        "○ ALGORITHMIC method: Alter loss function to incorporate more challenges. Use a\n",
        "multi-tasking approach to achieve your aim.Report the accuracy after mitigation and compare it with previous classification results."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hXnUIaktzwA6"
      },
      "source": [
        "dataset link: http://vis-www.cs.umass.edu/lfw/\n",
        "\n",
        "attributes link: https://www.cs.columbia.edu/CAVE/databases/pubfig/download/lfw_attributes.txt"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LAKUPchinboZ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9800b87c-6c8f-4c4b-a75c-82f1ddb7bff4"
      },
      "source": [
        "\n",
        "import numpy as np\n",
        "import os\n",
        "import cv2\n",
        "import pandas as pd\n",
        "import tarfile\n",
        "import tqdm\n",
        "\n",
        "# Google Drive access\n",
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')\n",
        "%cd /content/gdrive/My Drive/LFW\n",
        "!pwd"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/gdrive\n",
            "/content/gdrive/My Drive/LFW\n",
            "/content/gdrive/My Drive/LFW\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I81-1kgzcchX",
        "outputId": "11d0516c-0e67-4646-b5cb-38de595fa29d"
      },
      "source": [
        "# Dumping files to drive\n",
        "!wget http://www.cs.columbia.edu/CAVE/databases/pubfig/download/lfw_attributes.txt\n",
        "!wget http://vis-www.cs.umass.edu/lfw/lfw.tgz"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-12-29 07:07:28--  http://www.cs.columbia.edu/CAVE/databases/pubfig/download/lfw_attributes.txt\n",
            "Resolving www.cs.columbia.edu (www.cs.columbia.edu)... 128.59.11.206\n",
            "Connecting to www.cs.columbia.edu (www.cs.columbia.edu)|128.59.11.206|:80... connected.\n",
            "HTTP request sent, awaiting response... 301 Moved Permanently\n",
            "Location: https://www.cs.columbia.edu/CAVE/databases/pubfig/download/lfw_attributes.txt [following]\n",
            "--2020-12-29 07:07:28--  https://www.cs.columbia.edu/CAVE/databases/pubfig/download/lfw_attributes.txt\n",
            "Connecting to www.cs.columbia.edu (www.cs.columbia.edu)|128.59.11.206|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 14879205 (14M) [text/plain]\n",
            "Saving to: ‘lfw_attributes.txt.8’\n",
            "\n",
            "lfw_attributes.txt. 100%[===================>]  14.19M  16.8MB/s    in 0.8s    \n",
            "\n",
            "2020-12-29 07:07:29 (16.8 MB/s) - ‘lfw_attributes.txt.8’ saved [14879205/14879205]\n",
            "\n",
            "--2020-12-29 07:07:29--  http://vis-www.cs.umass.edu/lfw/lfw.tgz\n",
            "Resolving vis-www.cs.umass.edu (vis-www.cs.umass.edu)... 128.119.244.95\n",
            "Connecting to vis-www.cs.umass.edu (vis-www.cs.umass.edu)|128.119.244.95|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 180566744 (172M) [application/x-gzip]\n",
            "Saving to: ‘lfw.tgz.8’\n",
            "\n",
            "lfw.tgz.8           100%[===================>] 172.20M  51.6MB/s    in 4.7s    \n",
            "\n",
            "2020-12-29 07:07:34 (36.5 MB/s) - ‘lfw.tgz.8’ saved [180566744/180566744]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JRWq2WqFdMqG"
      },
      "source": [
        "# File handlers\n",
        "attributes = \"lfw_attributes.txt\"\n",
        "images = \"lfw.tgz\""
      ],
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "76P8wvhJdwu_"
      },
      "source": [
        "# ref: https://colab.research.google.com/github/MalihaUCF/Autoencoders/blob/master/Autoencoder_PCA1.ipynb#scrollTo=C0miIeXh9dy1&line=28&uniqifier=1\n",
        "\n",
        "def decode_image_from_raw_bytes(raw_bytes):\n",
        "    img = cv2.imdecode(np.asarray(bytearray(raw_bytes), dtype=np.uint8), 1)\n",
        "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n",
        "    return img\n",
        "\n",
        "\n",
        "def load_lfw_dataset(\n",
        "        use_raw=True,\n",
        "        dx=80, dy=80,\n",
        "        dimx=45, dimy=45):\n",
        "\n",
        "    # read attrs\n",
        "    df_attrs = pd.read_csv(attributes, sep='\\t', skiprows=1)\n",
        "    df_attrs = pd.DataFrame(df_attrs.iloc[:, :-1].values, columns=df_attrs.columns[1:])\n",
        "    imgs_with_attrs = set(map(tuple, df_attrs[[\"person\", \"imagenum\"]].values))\n",
        "\n",
        "    # read photos\n",
        "    all_photos = []\n",
        "    photo_ids = []\n",
        "\n",
        "    with tarfile.open(images) as f:\n",
        "        for m in tqdm.tqdm_notebook(f.getmembers()):\n",
        "            if m.isfile() and m.name.endswith(\".jpg\"):\n",
        "                # prepare image\n",
        "                img = decode_image_from_raw_bytes(f.extractfile(m).read())\n",
        "                img = img[dy:-dy, dx:-dx]\n",
        "                img = cv2.resize(img, (dimx, dimy))\n",
        "                # parse person\n",
        "                fname = os.path.split(m.name)[-1]\n",
        "                fname_splitted = fname[:-4].replace('_', ' ').split()\n",
        "                person_id = ' '.join(fname_splitted[:-1])\n",
        "                photo_number = int(fname_splitted[-1])\n",
        "                if (person_id, photo_number) in imgs_with_attrs:\n",
        "                    all_photos.append(img)\n",
        "                    photo_ids.append({'person': person_id, 'imagenum': photo_number})\n",
        "\n",
        "    photo_ids = pd.DataFrame(photo_ids)\n",
        "    all_photos = np.stack(all_photos).astype('uint8')\n",
        "\n",
        "    # preserve photo_ids order!\n",
        "    all_attrs = photo_ids.merge(df_attrs, on=('person', 'imagenum')).drop([\"person\", \"imagenum\"], axis=1)\n",
        "\n",
        "    return all_photos, all_attrs\n"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 385,
          "referenced_widgets": [
            "5b95b5621457464daaab970dc0158d5a",
            "b221d97ce2004f1388ea7a507232015b",
            "28032979b44043e3b6a3e3668791da3c",
            "d715ccd4fa2349079418a9498c943638",
            "15fb97ef90934fd4bbcb271315f44b9c",
            "a34a6b326d3e46e9b02cf2afd7387e4e",
            "6ec468aad5914fba9d7acc0cc0c696c2",
            "7e49534410ce4de09893103761712024"
          ]
        },
        "id": "HT3TefhNhDbW",
        "outputId": "6497e830-8bc2-4448-d6a7-f0d59375b44a"
      },
      "source": [
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "#from lfw_dataset import load_lfw_dataset\n",
        "import tensorflow as tf\n",
        "import keras, keras.layers as L\n",
        "\n",
        "# Loading and normalizing [Might take some time]\n",
        "X, attributes = load_lfw_dataset(use_raw=True,dimx=38,dimy=38)\n",
        "X = X.astype('float32') / 255.0\n",
        "img_shape = X.shape[1:]\n",
        "X_train, X_test = train_test_split(X, test_size=0.1, random_state=42)\n",
        "# Checking out some images\n",
        "%matplotlib inline\n",
        "import matplotlib.pyplot as plt\n",
        "plt.rcParams['axes.grid'] = False\n",
        "plt.title('sample image')\n",
        "for i in range(6):\n",
        "    plt.subplot(2,3,i+1)\n",
        "    plt.imshow(X[i])\n",
        "print(\"X shape:\",X.shape)\n",
        "print(\"attr shape:\",attributes.shape)"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:24: TqdmDeprecationWarning: This function will be removed in tqdm==5.0.0\n",
            "Please use `tqdm.notebook.tqdm` instead of `tqdm.tqdm_notebook`\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "5b95b5621457464daaab970dc0158d5a",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=18983.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "X shape: (13143, 38, 38, 3)\n",
            "attr shape: (13143, 73)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 6 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nYkV1eSIihrr"
      },
      "source": [
        "# lets divide the dataset into 65% and 35%"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i5-5Q0OoWwNY"
      },
      "source": [
        "X_95 = np.round(0.95*len(X))"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "53-g08ND-9SC",
        "outputId": "a19fcfdd-f056-4c55-9887-0b6d43035c57"
      },
      "source": [
        "X_95"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "12486.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RviIkbdk_F5Y",
        "outputId": "95adacad-9444-4531-a4fe-63a70d36599c"
      },
      "source": [
        "attributes.Eyeglasses.iloc[0]\n",
        "attributes['Eyeglasses_Sunglasses'] = 0\n",
        "for i in range(len(attributes)):\n",
        "  # to consider Eyeglasses and Sunglases as same quantity, let us take average of both and add it to the dataframe\n",
        "  if(np.sign(attributes.Eyeglasses.iloc[i])==1 or np.sign(attributes.Sunglasses.iloc[i])==1):\n",
        "    attributes['Eyeglasses_Sunglasses'].iloc[i] = 1\n",
        "  else:\n",
        "    attributes['Eyeglasses_Sunglasses'].iloc[i] = 0\n"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/pandas/core/indexing.py:670: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  iloc._setitem_with_indexer(indexer, value)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 275
        },
        "id": "_GSr1HmXDUwI",
        "outputId": "c36429ce-3101-4ae5-85fd-786b799794c9"
      },
      "source": [
        "attributes.head()"
      ],
      "execution_count": 27,
      "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>Male</th>\n",
              "      <th>Asian</th>\n",
              "      <th>White</th>\n",
              "      <th>Black</th>\n",
              "      <th>Baby</th>\n",
              "      <th>Child</th>\n",
              "      <th>Youth</th>\n",
              "      <th>Middle Aged</th>\n",
              "      <th>Senior</th>\n",
              "      <th>Black Hair</th>\n",
              "      <th>Blond Hair</th>\n",
              "      <th>Brown Hair</th>\n",
              "      <th>Bald</th>\n",
              "      <th>No Eyewear</th>\n",
              "      <th>Eyeglasses</th>\n",
              "      <th>Sunglasses</th>\n",
              "      <th>Mustache</th>\n",
              "      <th>Smiling</th>\n",
              "      <th>Frowning</th>\n",
              "      <th>Chubby</th>\n",
              "      <th>Blurry</th>\n",
              "      <th>Harsh Lighting</th>\n",
              "      <th>Flash</th>\n",
              "      <th>Soft Lighting</th>\n",
              "      <th>Outdoor</th>\n",
              "      <th>Curly Hair</th>\n",
              "      <th>Wavy Hair</th>\n",
              "      <th>Straight Hair</th>\n",
              "      <th>Receding Hairline</th>\n",
              "      <th>Bangs</th>\n",
              "      <th>Sideburns</th>\n",
              "      <th>Fully Visible Forehead</th>\n",
              "      <th>Partially Visible Forehead</th>\n",
              "      <th>Obstructed Forehead</th>\n",
              "      <th>Bushy Eyebrows</th>\n",
              "      <th>Arched Eyebrows</th>\n",
              "      <th>Narrow Eyes</th>\n",
              "      <th>Eyes Open</th>\n",
              "      <th>Big Nose</th>\n",
              "      <th>Pointy Nose</th>\n",
              "      <th>Big Lips</th>\n",
              "      <th>Mouth Closed</th>\n",
              "      <th>Mouth Slightly Open</th>\n",
              "      <th>Mouth Wide Open</th>\n",
              "      <th>Teeth Not Visible</th>\n",
              "      <th>No Beard</th>\n",
              "      <th>Goatee</th>\n",
              "      <th>Round Jaw</th>\n",
              "      <th>Double Chin</th>\n",
              "      <th>Wearing Hat</th>\n",
              "      <th>Oval Face</th>\n",
              "      <th>Square Face</th>\n",
              "      <th>Round Face</th>\n",
              "      <th>Color Photo</th>\n",
              "      <th>Posed Photo</th>\n",
              "      <th>Attractive Man</th>\n",
              "      <th>Attractive Woman</th>\n",
              "      <th>Indian</th>\n",
              "      <th>Gray Hair</th>\n",
              "      <th>Bags Under Eyes</th>\n",
              "      <th>Heavy Makeup</th>\n",
              "      <th>Rosy Cheeks</th>\n",
              "      <th>Shiny Skin</th>\n",
              "      <th>Pale Skin</th>\n",
              "      <th>5 o' Clock Shadow</th>\n",
              "      <th>Strong Nose-Mouth Lines</th>\n",
              "      <th>Wearing Lipstick</th>\n",
              "      <th>Flushed Face</th>\n",
              "      <th>High Cheekbones</th>\n",
              "      <th>Brown Eyes</th>\n",
              "      <th>Wearing Earrings</th>\n",
              "      <th>Wearing Necktie</th>\n",
              "      <th>Wearing Necklace</th>\n",
              "      <th>Eyeglasses_Sunglasses</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-0.602662</td>\n",
              "      <td>-0.83636</td>\n",
              "      <td>0.909554</td>\n",
              "      <td>-2.03901</td>\n",
              "      <td>-0.937008</td>\n",
              "      <td>-0.520853</td>\n",
              "      <td>-0.399348</td>\n",
              "      <td>-0.609662</td>\n",
              "      <td>-0.328941</td>\n",
              "      <td>-1.20055</td>\n",
              "      <td>-0.39051</td>\n",
              "      <td>-2.73407</td>\n",
              "      <td>-1.51945</td>\n",
              "      <td>0.555006</td>\n",
              "      <td>-0.625682</td>\n",
              "      <td>-1.45926</td>\n",
              "      <td>-0.866737</td>\n",
              "      <td>-0.396046</td>\n",
              "      <td>0.551072</td>\n",
              "      <td>-0.986439</td>\n",
              "      <td>-0.0734218</td>\n",
              "      <td>-0.564658</td>\n",
              "      <td>-1.16134</td>\n",
              "      <td>1.57563</td>\n",
              "      <td>0.965746</td>\n",
              "      <td>-0.607315</td>\n",
              "      <td>0.44931</td>\n",
              "      <td>0.668191</td>\n",
              "      <td>-0.405056</td>\n",
              "      <td>-0.613886</td>\n",
              "      <td>-0.167452</td>\n",
              "      <td>-2.5441</td>\n",
              "      <td>-0.58652</td>\n",
              "      <td>0.460766</td>\n",
              "      <td>0.00928555</td>\n",
              "      <td>-0.830431</td>\n",
              "      <td>1.97026</td>\n",
              "      <td>0.488666</td>\n",
              "      <td>-0.0477713</td>\n",
              "      <td>0.303884</td>\n",
              "      <td>-1.35427</td>\n",
              "      <td>0.930372</td>\n",
              "      <td>-1.71259</td>\n",
              "      <td>-0.298748</td>\n",
              "      <td>0.465369</td>\n",
              "      <td>0.794791</td>\n",
              "      <td>-0.450567</td>\n",
              "      <td>0.283735</td>\n",
              "      <td>-0.148479</td>\n",
              "      <td>1.27906</td>\n",
              "      <td>0.959224</td>\n",
              "      <td>-0.7095</td>\n",
              "      <td>-0.684215</td>\n",
              "      <td>2.31964</td>\n",
              "      <td>-0.858775</td>\n",
              "      <td>0.0266835</td>\n",
              "      <td>-0.929809</td>\n",
              "      <td>-1.79466</td>\n",
              "      <td>0.094929</td>\n",
              "      <td>-0.56723</td>\n",
              "      <td>-1.07044</td>\n",
              "      <td>-1.52884</td>\n",
              "      <td>-1.4068</td>\n",
              "      <td>-0.793891</td>\n",
              "      <td>-1.77256</td>\n",
              "      <td>0.234426</td>\n",
              "      <td>-0.531727</td>\n",
              "      <td>-2.35169</td>\n",
              "      <td>0.0929558</td>\n",
              "      <td>1.51383</td>\n",
              "      <td>-0.176601</td>\n",
              "      <td>-0.335991</td>\n",
              "      <td>-0.882286</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.505597</td>\n",
              "      <td>-1.91405</td>\n",
              "      <td>2.51025</td>\n",
              "      <td>-2.4314</td>\n",
              "      <td>-1.00337</td>\n",
              "      <td>-1.44539</td>\n",
              "      <td>-1.537</td>\n",
              "      <td>-0.204474</td>\n",
              "      <td>0.487091</td>\n",
              "      <td>-1.74747</td>\n",
              "      <td>-0.311364</td>\n",
              "      <td>-0.0631772</td>\n",
              "      <td>-0.957594</td>\n",
              "      <td>0.501207</td>\n",
              "      <td>0.221517</td>\n",
              "      <td>-1.75726</td>\n",
              "      <td>-0.551377</td>\n",
              "      <td>-0.282309</td>\n",
              "      <td>0.195253</td>\n",
              "      <td>-0.317791</td>\n",
              "      <td>-0.912739</td>\n",
              "      <td>-0.614058</td>\n",
              "      <td>0.64591</td>\n",
              "      <td>0.481944</td>\n",
              "      <td>-1.51933</td>\n",
              "      <td>0.566179</td>\n",
              "      <td>0.801907</td>\n",
              "      <td>-1.0607</td>\n",
              "      <td>0.901279</td>\n",
              "      <td>-1.34742</td>\n",
              "      <td>0.0797163</td>\n",
              "      <td>0.268679</td>\n",
              "      <td>-0.467019</td>\n",
              "      <td>-1.30433</td>\n",
              "      <td>-0.991695</td>\n",
              "      <td>-0.0503162</td>\n",
              "      <td>0.261866</td>\n",
              "      <td>0.334746</td>\n",
              "      <td>0.785039</td>\n",
              "      <td>1.26984</td>\n",
              "      <td>-1.25441</td>\n",
              "      <td>1.10048</td>\n",
              "      <td>-0.924761</td>\n",
              "      <td>-1.37896</td>\n",
              "      <td>0.800159</td>\n",
              "      <td>0.343087</td>\n",
              "      <td>0.276153</td>\n",
              "      <td>0.115633</td>\n",
              "      <td>0.761935</td>\n",
              "      <td>-0.668475</td>\n",
              "      <td>0.721088</td>\n",
              "      <td>-0.480292</td>\n",
              "      <td>-1.00298</td>\n",
              "      <td>3.99115</td>\n",
              "      <td>-0.345703</td>\n",
              "      <td>0.0431644</td>\n",
              "      <td>-1.38426</td>\n",
              "      <td>-1.93805</td>\n",
              "      <td>0.39326</td>\n",
              "      <td>-0.749561</td>\n",
              "      <td>-0.329255</td>\n",
              "      <td>-0.579444</td>\n",
              "      <td>-0.234069</td>\n",
              "      <td>0.0096608</td>\n",
              "      <td>-0.374216</td>\n",
              "      <td>-0.0136298</td>\n",
              "      <td>-0.0809645</td>\n",
              "      <td>-2.12532</td>\n",
              "      <td>0.281169</td>\n",
              "      <td>-1.24167</td>\n",
              "      <td>-0.515579</td>\n",
              "      <td>0.1815</td>\n",
              "      <td>-0.80375</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.4038</td>\n",
              "      <td>-1.39616</td>\n",
              "      <td>0.636715</td>\n",
              "      <td>-1.02491</td>\n",
              "      <td>-1.76605</td>\n",
              "      <td>-0.661043</td>\n",
              "      <td>-1.73453</td>\n",
              "      <td>-0.239029</td>\n",
              "      <td>0.350139</td>\n",
              "      <td>-1.67061</td>\n",
              "      <td>-0.239739</td>\n",
              "      <td>-1.02587</td>\n",
              "      <td>-0.669365</td>\n",
              "      <td>0.294301</td>\n",
              "      <td>-0.243044</td>\n",
              "      <td>-1.24041</td>\n",
              "      <td>-0.559538</td>\n",
              "      <td>0.900781</td>\n",
              "      <td>-0.864989</td>\n",
              "      <td>0.925484</td>\n",
              "      <td>-0.415565</td>\n",
              "      <td>0.477309</td>\n",
              "      <td>-0.963641</td>\n",
              "      <td>0.498012</td>\n",
              "      <td>0.737615</td>\n",
              "      <td>1.38272</td>\n",
              "      <td>-0.318062</td>\n",
              "      <td>-0.137686</td>\n",
              "      <td>0.824265</td>\n",
              "      <td>-0.209637</td>\n",
              "      <td>-0.138066</td>\n",
              "      <td>0.0272288</td>\n",
              "      <td>-0.4622</td>\n",
              "      <td>-0.283984</td>\n",
              "      <td>-0.0591731</td>\n",
              "      <td>-0.364826</td>\n",
              "      <td>1.32351</td>\n",
              "      <td>0.275434</td>\n",
              "      <td>1.30362</td>\n",
              "      <td>-0.126634</td>\n",
              "      <td>-0.596104</td>\n",
              "      <td>-0.337977</td>\n",
              "      <td>-0.486416</td>\n",
              "      <td>0.0556214</td>\n",
              "      <td>-0.177267</td>\n",
              "      <td>0.375264</td>\n",
              "      <td>-0.448152</td>\n",
              "      <td>0.211282</td>\n",
              "      <td>0.64118</td>\n",
              "      <td>-1.23075</td>\n",
              "      <td>-1.00435</td>\n",
              "      <td>-0.264218</td>\n",
              "      <td>-0.173436</td>\n",
              "      <td>0.983262</td>\n",
              "      <td>-1.02405</td>\n",
              "      <td>-0.968107</td>\n",
              "      <td>-1.25158</td>\n",
              "      <td>-0.831154</td>\n",
              "      <td>-0.0487521</td>\n",
              "      <td>1.87388</td>\n",
              "      <td>-1.28517</td>\n",
              "      <td>0.366453</td>\n",
              "      <td>-0.517039</td>\n",
              "      <td>1.11466</td>\n",
              "      <td>-0.503967</td>\n",
              "      <td>1.65218</td>\n",
              "      <td>-1.14525</td>\n",
              "      <td>-1.25307</td>\n",
              "      <td>0.874682</td>\n",
              "      <td>0.497179</td>\n",
              "      <td>-1.11823</td>\n",
              "      <td>0.719109</td>\n",
              "      <td>-0.163028</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.929768</td>\n",
              "      <td>-1.75116</td>\n",
              "      <td>2.34706</td>\n",
              "      <td>-2.24259</td>\n",
              "      <td>-1.35852</td>\n",
              "      <td>-0.902918</td>\n",
              "      <td>-1.87211</td>\n",
              "      <td>-0.348193</td>\n",
              "      <td>0.106553</td>\n",
              "      <td>-1.92001</td>\n",
              "      <td>-0.843878</td>\n",
              "      <td>-1.45527</td>\n",
              "      <td>0.116413</td>\n",
              "      <td>-0.00884933</td>\n",
              "      <td>0.0968053</td>\n",
              "      <td>-1.77056</td>\n",
              "      <td>-0.883169</td>\n",
              "      <td>0.251477</td>\n",
              "      <td>-0.222816</td>\n",
              "      <td>0.979753</td>\n",
              "      <td>-0.485323</td>\n",
              "      <td>-1.15441</td>\n",
              "      <td>-0.31113</td>\n",
              "      <td>-0.157736</td>\n",
              "      <td>-0.350956</td>\n",
              "      <td>0.950488</td>\n",
              "      <td>0.347889</td>\n",
              "      <td>-1.20252</td>\n",
              "      <td>1.60015</td>\n",
              "      <td>-0.130796</td>\n",
              "      <td>-0.173291</td>\n",
              "      <td>-0.149002</td>\n",
              "      <td>-0.510854</td>\n",
              "      <td>-0.382419</td>\n",
              "      <td>-1.09431</td>\n",
              "      <td>-0.205277</td>\n",
              "      <td>0.963489</td>\n",
              "      <td>0.853781</td>\n",
              "      <td>0.156086</td>\n",
              "      <td>0.312227</td>\n",
              "      <td>-1.82151</td>\n",
              "      <td>0.594445</td>\n",
              "      <td>-0.243877</td>\n",
              "      <td>-1.54746</td>\n",
              "      <td>0.0892587</td>\n",
              "      <td>0.294815</td>\n",
              "      <td>-0.338663</td>\n",
              "      <td>0.0352336</td>\n",
              "      <td>1.02599</td>\n",
              "      <td>-0.407141</td>\n",
              "      <td>-0.64235</td>\n",
              "      <td>-0.298058</td>\n",
              "      <td>-0.607125</td>\n",
              "      <td>2.074</td>\n",
              "      <td>-0.076147</td>\n",
              "      <td>-0.79068</td>\n",
              "      <td>-1.32389</td>\n",
              "      <td>-1.5935</td>\n",
              "      <td>0.405241</td>\n",
              "      <td>0.768172</td>\n",
              "      <td>-1.27286</td>\n",
              "      <td>-1.30234</td>\n",
              "      <td>-0.816332</td>\n",
              "      <td>1.7813</td>\n",
              "      <td>-1.03106</td>\n",
              "      <td>0.705507</td>\n",
              "      <td>-1.68246</td>\n",
              "      <td>0.0304102</td>\n",
              "      <td>0.303866</td>\n",
              "      <td>-0.996759</td>\n",
              "      <td>-0.562108</td>\n",
              "      <td>1.39505</td>\n",
              "      <td>-0.601105</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.38212</td>\n",
              "      <td>-1.83819</td>\n",
              "      <td>1.63398</td>\n",
              "      <td>-1.49223</td>\n",
              "      <td>-1.35611</td>\n",
              "      <td>-0.560854</td>\n",
              "      <td>-1.68384</td>\n",
              "      <td>-0.825507</td>\n",
              "      <td>0.886115</td>\n",
              "      <td>-0.712738</td>\n",
              "      <td>-0.661126</td>\n",
              "      <td>-1.24526</td>\n",
              "      <td>-0.791411</td>\n",
              "      <td>1.20355</td>\n",
              "      <td>0.0238365</td>\n",
              "      <td>-0.901334</td>\n",
              "      <td>-0.807165</td>\n",
              "      <td>-1.22825</td>\n",
              "      <td>1.55971</td>\n",
              "      <td>0.280109</td>\n",
              "      <td>-1.60343</td>\n",
              "      <td>0.0921409</td>\n",
              "      <td>-0.770683</td>\n",
              "      <td>1.2718</td>\n",
              "      <td>0.0497857</td>\n",
              "      <td>0.109589</td>\n",
              "      <td>-0.910338</td>\n",
              "      <td>-1.046</td>\n",
              "      <td>-0.14786</td>\n",
              "      <td>-0.723328</td>\n",
              "      <td>-0.1958</td>\n",
              "      <td>0.970945</td>\n",
              "      <td>0.0858819</td>\n",
              "      <td>-1.36246</td>\n",
              "      <td>-0.221446</td>\n",
              "      <td>-1.39699</td>\n",
              "      <td>2.25194</td>\n",
              "      <td>0.369097</td>\n",
              "      <td>0.498237</td>\n",
              "      <td>0.365376</td>\n",
              "      <td>-1.93622</td>\n",
              "      <td>0.00612372</td>\n",
              "      <td>-0.695058</td>\n",
              "      <td>-1.05815</td>\n",
              "      <td>0.45623</td>\n",
              "      <td>0.344706</td>\n",
              "      <td>-0.441983</td>\n",
              "      <td>0.217264</td>\n",
              "      <td>0.237759</td>\n",
              "      <td>-1.22728</td>\n",
              "      <td>0.0431653</td>\n",
              "      <td>-0.166989</td>\n",
              "      <td>-0.155174</td>\n",
              "      <td>-7.9169</td>\n",
              "      <td>0.522702</td>\n",
              "      <td>-1.1373</td>\n",
              "      <td>-0.845506</td>\n",
              "      <td>-1.40812</td>\n",
              "      <td>0.0270361</td>\n",
              "      <td>0.24311</td>\n",
              "      <td>-0.688967</td>\n",
              "      <td>-7.0227</td>\n",
              "      <td>-1.04577</td>\n",
              "      <td>3.26579</td>\n",
              "      <td>-1.12066</td>\n",
              "      <td>0.374019</td>\n",
              "      <td>-1.55722</td>\n",
              "      <td>-11.8743</td>\n",
              "      <td>-0.479549</td>\n",
              "      <td>-0.990555</td>\n",
              "      <td>-0.655915</td>\n",
              "      <td>0.912489</td>\n",
              "      <td>-0.308988</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       Male    Asian  ... Wearing Necklace Eyeglasses_Sunglasses\n",
              "0 -0.602662 -0.83636  ...        -0.882286                     0\n",
              "1  0.505597 -1.91405  ...         -0.80375                     1\n",
              "2    1.4038 -1.39616  ...        -0.163028                     0\n",
              "3  0.929768 -1.75116  ...        -0.601105                     1\n",
              "4   1.38212 -1.83819  ...        -0.308988                     1\n",
              "\n",
              "[5 rows x 74 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "slRp1oYX_2hY",
        "outputId": "821bf372-2375-445d-985d-e25663795bd4"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.imshow(X[0])"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f7205044c88>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VUF-xybDaQ6G"
      },
      "source": [
        "Xtrain and Ytrain\n",
        "\n",
        "Xtest and Ytest"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZnB88N7ux2ae"
      },
      "source": [
        "# Start from here : The mitigation part. Taking 95% data as training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "je_40aWpB0QK",
        "outputId": "04b0635c-a24c-4289-c008-4c22b18e71fc"
      },
      "source": [
        "X_train_95 = X[0:12486]\n",
        "len(X_train_95)\n",
        "\n",
        "X_train_95 = X_train_95[0:12486]\n",
        "X_test_95 = X[12486:13143]\n",
        "\n",
        "print(len(X_train_95))\n",
        "print(len(X_test_95))\n",
        "\n",
        "Xtrain = X_train_95\n",
        "Xtest = X_test_95\n",
        "Ydata = np.array(attributes.Eyeglasses_Sunglasses)\n",
        "print(len(Ydata))\n",
        "print(Ydata)\n",
        "Y_95 = Ydata[0:13143]\n",
        "Y_train_95 = Y_95[0:12486] # Taking 80% data\n",
        "Y_test_95 = Y_95[12486:13143]\n",
        "print(len(Y_train_95))\n",
        "print(len(Y_test_95))\n",
        "Ytrain = Y_train_95\n",
        "Ytest = Y_test_95\n"
      ],
      "execution_count": 89,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "12486\n",
            "657\n",
            "13143\n",
            "[0 1 0 ... 0 0 0]\n",
            "12486\n",
            "657\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xrkdqVZODQdO",
        "outputId": "f1fb6868-faf7-4fb8-b380-56f352a8bfcb"
      },
      "source": [
        "Xtrain = Xtrain[:,:,:,0]\n",
        "Xtest = Xtest[:,:,:,0]\n",
        "print(Xtrain.shape)\n",
        "print(Ytrain.shape)\n",
        "print(Xtest.shape)\n",
        "print(Ytest.shape)"
      ],
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(12486, 38, 38)\n",
            "(12486,)\n",
            "(657, 38, 38)\n",
            "(657,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IRZSmBvXbqfE",
        "outputId": "0f74d50b-8619-45bd-c27b-09fc61e4d26c"
      },
      "source": [
        "print(type(Xtrain))\n",
        "print(type(Ytrain))\n",
        "print(type(Xtest))\n",
        "print(type(Ytest))"
      ],
      "execution_count": 91,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'numpy.ndarray'>\n",
            "<class 'numpy.ndarray'>\n",
            "<class 'numpy.ndarray'>\n",
            "<class 'numpy.ndarray'>\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y7rna6LRAONe",
        "outputId": "be23e27f-a76c-4514-b70f-96a0cb85619d"
      },
      "source": [
        "Xtrain = Xtrain.reshape(12486,38*38)\n",
        "Xtest = Xtest.reshape(657,38*38)\n",
        "print(Xtrain.shape)\n",
        "print(Ytrain.shape)\n",
        "print(Xtest.shape)\n",
        "print(Ytest.shape)"
      ],
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(12486, 1444)\n",
            "(12486,)\n",
            "(657, 1444)\n",
            "(657,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JtZGSUVle73F",
        "outputId": "a2d7d654-22cc-4d50-fe8e-26d285fd0536"
      },
      "source": [
        "from skimage import data\n",
        "from skimage.color import rgb2gray\n",
        "Xtrain_rgb2grey = rgb2gray(Xtrain)\n",
        "Xtest_rgb2grey = rgb2gray(Xtest)\n",
        "print(Xtrain_rgb2grey.shape)\n",
        "print(Ytrain.shape)\n",
        "print(Xtest_rgb2grey.shape)\n",
        "print(Ytest.shape)"
      ],
      "execution_count": 93,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(12486, 1444)\n",
            "(12486,)\n",
            "(657, 1444)\n",
            "(657,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EwDRq4VOiJje",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c608bca5-abed-424e-f693-adad2255289e"
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras import layers, models\n",
        "\n",
        "ann = models.Sequential([\n",
        "        # layers.Flatten(input_shape=(1444)),\n",
        "        layers.Dense(128, activation='relu'),\n",
        "        layers.Dense(128, activation='relu'),\n",
        "        layers.Dense(128, activation='relu'),\n",
        "        layers.Dense(64, activation='relu'),\n",
        "        layers.Dense(1, activation='sigmoid')    \n",
        "    ])\n",
        "\n",
        "ann.compile(optimizer='SGD',\n",
        "              loss='mean_squared_error',\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "ann.fit(Xtrain, Ytrain, epochs=20)"
      ],
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "391/391 [==============================] - 2s 3ms/step - loss: 0.1642 - accuracy: 0.7975\n",
            "Epoch 2/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1552 - accuracy: 0.8079\n",
            "Epoch 3/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1521 - accuracy: 0.8095\n",
            "Epoch 4/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1509 - accuracy: 0.8089\n",
            "Epoch 5/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1481 - accuracy: 0.8095\n",
            "Epoch 6/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1418 - accuracy: 0.8160\n",
            "Epoch 7/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1410 - accuracy: 0.8107\n",
            "Epoch 8/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1399 - accuracy: 0.8084\n",
            "Epoch 9/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1347 - accuracy: 0.8177\n",
            "Epoch 10/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1324 - accuracy: 0.8210\n",
            "Epoch 11/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1335 - accuracy: 0.8205\n",
            "Epoch 12/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1298 - accuracy: 0.8221\n",
            "Epoch 13/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1267 - accuracy: 0.8267\n",
            "Epoch 14/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1272 - accuracy: 0.8258\n",
            "Epoch 15/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1308 - accuracy: 0.8263\n",
            "Epoch 16/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1271 - accuracy: 0.8312\n",
            "Epoch 17/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1223 - accuracy: 0.8344\n",
            "Epoch 18/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1288 - accuracy: 0.8196\n",
            "Epoch 19/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1265 - accuracy: 0.8252\n",
            "Epoch 20/20\n",
            "391/391 [==============================] - 1s 3ms/step - loss: 0.1247 - accuracy: 0.8318\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f71f97f8240>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 94
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C8LCHAt-EAWm",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "outputId": "cb479921-81ef-43fb-c229-a967205de349"
      },
      "source": [
        "classes = ['0','1']\n",
        "from sklearn.metrics import confusion_matrix , classification_report, accuracy_score\n",
        "import numpy as np\n",
        "Ypred = ann.predict(Xtest)\n",
        "Ypred_classes = Ypred\n",
        "\n",
        "for i in range(len(Ypred)):\n",
        "  if(Ypred[i]>0.5):\n",
        "    Ypred_classes[i]=1\n",
        "  else:\n",
        "    Ypred_classes[i]=0\n",
        "# Ypred_classes = [np.argmax(element) for element in Ypred]\n",
        "\n",
        "print(\"Classification Report: \\n\", classification_report(Ytest, Ypred_classes))\n",
        "print(\"-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\")\n",
        "print(\"Confusion Matrix: \\n\", confusion_matrix(Ytest, Ypred_classes))\n",
        "print(\"-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\")\n",
        "print(\"Accuracy Score: \\n\", accuracy_score(Ytest, Ypred_classes))\n",
        "print(\"-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\")\n",
        "print(\"along x axis - true, along y axis - predicted\")\n",
        "\n",
        "import seaborn as sns\n",
        "\n",
        "matrix = confusion_matrix(Ytest, Ypred_classes)\n",
        "sns.heatmap(matrix, square=True, annot=True, fmt='d', cbar=False,\n",
        "            xticklabels=classes,\n",
        "            yticklabels=classes)\n",
        "plt.xlabel('true label')\n",
        "plt.ylabel('predicted label');"
      ],
      "execution_count": 95,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report: \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "           0       0.86      1.00      0.92       559\n",
            "           1       0.78      0.07      0.13        98\n",
            "\n",
            "    accuracy                           0.86       657\n",
            "   macro avg       0.82      0.53      0.53       657\n",
            "weighted avg       0.85      0.86      0.80       657\n",
            "\n",
            "-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\n",
            "Confusion Matrix: \n",
            " [[557   2]\n",
            " [ 91   7]]\n",
            "-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\n",
            "Accuracy Score: \n",
            " 0.8584474885844748\n",
            "-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\n",
            "along x axis - true, along y axis - predicted\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 602
        },
        "id": "Mz9_oxSgZSKp",
        "outputId": "6e396741-89f6-4d64-b6a5-ffd88c6b1c0a"
      },
      "source": [
        "from scipy.special import softmax\n",
        "softmaxes_prediction = softmax(Ypred)\n",
        "softmax_classes = softmaxes_prediction\n",
        "\n",
        "for i in range(len(Ypred)):\n",
        "  if(Ypred[i]>0.5):\n",
        "    softmax_classes[i]=1\n",
        "  else:\n",
        "    softmax_classes[i]=0\n",
        "\n",
        "\n",
        "print(\"Classification Report: \\n\", classification_report(Ytest, softmax_classes))\n",
        "print(\"-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\")\n",
        "print(\"Confusion Matrix: \\n\", confusion_matrix(Ytest, softmax_classes))\n",
        "print(\"-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\")\n",
        "print(\"Accuracy Score: \\n\", accuracy_score(Ytest, softmax_classes))\n",
        "print(\"-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\")\n",
        "print(\"along x axis - true, along y axis - predicted\")\n",
        "\n",
        "import seaborn as sns\n",
        "\n",
        "matrix = confusion_matrix(Ytest, softmax_classes)\n",
        "sns.heatmap(matrix, square=True, annot=True, fmt='d', cbar=False,\n",
        "            xticklabels=classes,\n",
        "            yticklabels=classes)\n",
        "plt.xlabel('true label')\n",
        "plt.ylabel('predicted label');"
      ],
      "execution_count": 103,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Classification Report: \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "           0       0.86      1.00      0.92       559\n",
            "           1       0.78      0.07      0.13        98\n",
            "\n",
            "    accuracy                           0.86       657\n",
            "   macro avg       0.82      0.53      0.53       657\n",
            "weighted avg       0.85      0.86      0.80       657\n",
            "\n",
            "-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\n",
            "Confusion Matrix: \n",
            " [[557   2]\n",
            " [ 91   7]]\n",
            "-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\n",
            "Accuracy Score: \n",
            " 0.8584474885844748\n",
            "-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\n",
            "along x axis - true, along y axis - predicted\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y0zWYDgLcZhN"
      },
      "source": [
        "def new_eval_metric_averaging(Ytest,Ypred_classes):\n",
        "  final_matrix = np.zeros((2,2))\n",
        "  final_matrix[0,0]=(matrix[0,0])/matrix[0,0]+matrix[0,1]\n",
        "  final_matrix[0,1]=(matrix[0,1])/matrix[0,0]+matrix[0,1]\n",
        "  final_matrix[1,0]=(matrix[1,0])/matrix[1,1]+matrix[1,0]\n",
        "  final_matrix[1,1]=(matrix[1,1])/matrix[1,1]+matrix[1,0]\n",
        "  return final_matrix.astype(int)\n",
        "# final_matrix[0,0]=0\n",
        "# final_matrix[0,1]=1\n",
        "# final_matrix[1,0]=2\n",
        "# final_matrix[1,1]=3\n",
        "# final_matrix"
      ],
      "execution_count": 122,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PLGU5MTDcsWI",
        "outputId": "17955c3a-e7d1-4f51-c03a-6939b849a166"
      },
      "source": [
        "print(\"-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\")\n",
        "print(\"New Eval Metric Matrix Averaging: \\n\", new_eval_metric_averaging(Ytest, Ypred_classes))\n"
      ],
      "execution_count": 123,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/-/\n",
            "New Eval Metric Matrix Averaging: \n",
            " [[  3   2]\n",
            " [104  92]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "00TnR6Medn6B"
      },
      "source": [
        ""
      ],
      "execution_count": 99,
      "outputs": []
    }
  ]
}