alzheimer_image_classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "alzheimer image classification",
      "provenance": [],
      "authorship_tag": "ABX9TyPW6PfsWxgKWoR4NmKR03bH",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/fadillarizalul/thesis-alzheimer/blob/main/alzheimer_image_classification.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# set up dan inisialisasi"
      ],
      "metadata": {
        "id": "8dVYnyo1hObz"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9BuCcYouVFc0",
        "outputId": "f97bbb32-6cbc-4faa-c705-e4fabd238a6f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "%matplotlib inline\n",
        "import seaborn as sns\n",
        "\n",
        "#import data dari drive\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!wget --no-check-certificate \\\n",
        "https://github.com/fadillarizalul/thesis-alzheimer/blob/main/dataset/3-classes-alzheimer-20-patients.zip?raw=true \\\n",
        "  -O 3-classes-alzheimer.zip"
      ],
      "metadata": {
        "id": "Jvso5lHEYymU",
        "outputId": "ae10ce71-ed94-41da-9446-ca2ed243098b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2022-09-11 12:58:10--  https://github.com/fadillarizalul/thesis-alzheimer/blob/main/dataset/3-classes-alzheimer-20-patients.zip?raw=true\n",
            "Resolving github.com (github.com)... 20.205.243.166\n",
            "Connecting to github.com (github.com)|20.205.243.166|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://github.com/fadillarizalul/thesis-alzheimer/raw/main/dataset/3-classes-alzheimer-20-patients.zip [following]\n",
            "--2022-09-11 12:58:10--  https://github.com/fadillarizalul/thesis-alzheimer/raw/main/dataset/3-classes-alzheimer-20-patients.zip\n",
            "Reusing existing connection to github.com:443.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://raw.githubusercontent.com/fadillarizalul/thesis-alzheimer/main/dataset/3-classes-alzheimer-20-patients.zip [following]\n",
            "--2022-09-11 12:58:11--  https://raw.githubusercontent.com/fadillarizalul/thesis-alzheimer/main/dataset/3-classes-alzheimer-20-patients.zip\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 9859577 (9.4M) [application/zip]\n",
            "Saving to: ‘3-classes-alzheimer.zip’\n",
            "\n",
            "3-classes-alzheimer 100%[===================>]   9.40M  --.-KB/s    in 0.03s   \n",
            "\n",
            "2022-09-11 12:58:13 (284 MB/s) - ‘3-classes-alzheimer.zip’ saved [9859577/9859577]\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Extracting ZIP files"
      ],
      "metadata": {
        "id": "ji6N-io5JTYS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!unzip /content/3-classes-alzheimer.zip"
      ],
      "metadata": {
        "id": "W-WJ4DSAJV0A",
        "outputId": "9db404b8-7ef0-4d0c-a2e7-44e48775e089",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Archive:  /content/3-classes-alzheimer.zip\n",
            "   creating: 3-classes-alzheimer-20-patients/\n",
            "   creating: 3-classes-alzheimer-20-patients/AD/\n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00001-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00001-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00001-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00001-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00001-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00002-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00002-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00002-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00002-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00002-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00003-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00003-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00003-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00003-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00003-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00006.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00007.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00008.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00004-00009.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00006.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00007.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00008.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00005-00009.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00006-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00006-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00006-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00006-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00006-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00007-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00007-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00007-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00007-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00007-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00008-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00008-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00008-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00008-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00008-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00009-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00009-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00009-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00009-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00009-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00010-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00010-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00010-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00010-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00010-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00011-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00011-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00011-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00011-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00011-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00012-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00012-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00012-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00012-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00012-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00013-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00013-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00013-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00013-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00013-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00014-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00014-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00014-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00014-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00014-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00015-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00015-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00015-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00015-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00015-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00016-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00016-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00016-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00016-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00016-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00017-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00017-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00017-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00017-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00017-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00018-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00018-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00018-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00018-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00018-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00019-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00019-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00019-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00019-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00019-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00020-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00020-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00020-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00020-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/AD/img-00020-00005.jpg  \n",
            "   creating: 3-classes-alzheimer-20-patients/CN/\n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00001-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00001-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00001-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00001-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00001-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00002-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00002-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00002-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00002-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00002-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00003-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00003-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00003-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00003-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00003-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00004-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00004-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00004-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00004-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00004-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00004-00006.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00005-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00005-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00005-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00005-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00005-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00006-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00006-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00006-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00006-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00006-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00007-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00007-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00007-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00007-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00007-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00008-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00008-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00008-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00008-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00008-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00009-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00009-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00009-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00009-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00009-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00010-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00010-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00010-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00010-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00010-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00011-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00011-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00011-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00011-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00011-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00012-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00012-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00012-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00012-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00012-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00013-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00013-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00013-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00013-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00013-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00014-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00014-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00014-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00014-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00014-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00015-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00015-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00015-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00015-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00015-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00016-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00016-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00016-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00016-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00016-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00017-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00017-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00017-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00017-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00017-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00018-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00018-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00018-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00018-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00018-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00019-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00019-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00019-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00019-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00019-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00020-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00020-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00020-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00020-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/CN/img-00020-00005.jpg  \n",
            "   creating: 3-classes-alzheimer-20-patients/MCI/\n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00001-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00001-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00001-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00001-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00001-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00002-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00002-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00002-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00002-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00002-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00003-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00003-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00003-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00003-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00003-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00004-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00004-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00004-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00004-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00004-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00005-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00005-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00005-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00005-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00005-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00006-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00006-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00006-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00006-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00006-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00007-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00007-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00007-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00007-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00007-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00008-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00008-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00008-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00008-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00008-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00009-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00009-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00009-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00009-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00009-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00010-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00010-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00010-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00010-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00010-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00011-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00011-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00011-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00011-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00011-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00012-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00012-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00012-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00012-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00012-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00013-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00013-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00013-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00013-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00013-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00014-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00014-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00014-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00014-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00014-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00015-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00015-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00015-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00015-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00015-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00016-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00016-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00016-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00016-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00016-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00017-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00017-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00017-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00017-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00017-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00018-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00018-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00018-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00018-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00018-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00019-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00019-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00019-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00019-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00019-00005.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00020-00001.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00020-00002.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00020-00003.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00020-00004.jpg  \n",
            "  inflating: 3-classes-alzheimer-20-patients/MCI/img-00020-00005.jpg  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Create base directory for each classes"
      ],
      "metadata": {
        "id": "KUx68G8HJgpG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        " #lokasi direktori dasar\n",
        "dir_base ='/content/3-classes-alzheimer-20-patients'\n",
        "os.listdir(dir_base)\n",
        "print(os.listdir(dir_base))\n",
        "\n",
        "#hitung jumlah file per direktori\n",
        "data_cn = len(os.listdir('/content/3-classes-alzheimer-20-patients/CN'))\n",
        "data_mci = len(os.listdir('/content/3-classes-alzheimer-20-patients/MCI')) \n",
        "data_ad = len(os.listdir('/content/3-classes-alzheimer-20-patients/AD')) \n",
        "\n",
        "print(\"Amount of CN class images:\", data_cn)\n",
        "print(\"Amount of MCI class images:\", data_mci)\n",
        "print(\"Amount of AD class images:\", data_ad)"
      ],
      "metadata": {
        "id": "gOiuaqGVJljF",
        "outputId": "f78db25d-08c9-46b3-f669-c22dd75b6eb9",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['CN', 'AD', 'MCI']\n",
            "Amount of CN class images: 101\n",
            "Amount of MCI class images: 100\n",
            "Amount of AD class images: 108\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# augmentasi dan split data"
      ],
      "metadata": {
        "id": "MPnipObYhSo1"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "image data generator dan split data dengan 80 20"
      ],
      "metadata": {
        "id": "CqC2LdmmKI_v"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "import keras_preprocessing\n",
        "from keras_preprocessing import image\n",
        "from keras_preprocessing.image import ImageDataGenerator"
      ],
      "metadata": {
        "id": "E0ybKmB2aKfu"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_datagen = ImageDataGenerator(rescale = 1./255,\n",
        "                                  # rotation_range=30,\n",
        "                                  # zoom_range=0.2,\n",
        "                                  # horizontal_flip=True,\n",
        "                                  # vertical_flip=True,\n",
        "                                   validation_split = 0.2\n",
        "                                   )\n",
        "\n",
        "# valid_datagen = ImageDataGenerator(rescale = 1./255,\n",
        "#                                    validation_split = 0.2)\n",
        "\n",
        "# test_datagen  = ImageDataGenerator(rescale = 1./255)"
      ],
      "metadata": {
        "id": "UvJGSRQfRT3d"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_generator = train_datagen.flow_from_directory(\n",
        "    directory = dir_base,\n",
        "    target_size = (256,256),\n",
        "    batch_size = 16,\n",
        "    class_mode = 'categorical',\n",
        "    subset='training'\n",
        ")\n",
        " \n",
        "val_generator = train_datagen.flow_from_directory(\n",
        "    directory = dir_base,\n",
        "    target_size = (256,256),\n",
        "    batch_size = 16,\n",
        "    class_mode = 'categorical',\n",
        "    subset='validation'\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w7MqWW2Uhdal",
        "outputId": "03672768-2a45-43bf-e5ab-2bd79791fed7"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 248 images belonging to 3 classes.\n",
            "Found 61 images belonging to 3 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Showing augmented data"
      ],
      "metadata": {
        "id": "ucX1BQ99wEKb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# from keras.preprocessing.image import ImageDataGenerator, array_to_img, img_to_array, load_img\n",
        "\n",
        "# datagen = ImageDataGenerator(\n",
        "#                             # rotation_range=90\n",
        "#                               # width_shift_range=0.3\n",
        "#                               #height_shift_range=0.3\n",
        "#                              # brightness_range=(0.5, 1.5)\n",
        "#                               # shear_range=0.2\n",
        "#                             # zoom_range=0.2\n",
        "#                             # horizontal_flip=True\n",
        "#                               # vertical_flip=True\n",
        "#                              )\n",
        "\n",
        "# img = load_img('/content/3-classes-alzheimer-20-patients/AD/img-00005-00008.jpg')  # this is a PIL image\n",
        "# x = img_to_array(img)  # this is a Numpy array with shape (3, 150, 150)\n",
        "# x = x.reshape((1,) + x.shape)  # this is a Numpy array with shape (1, 3, 150, 150)\n",
        "\n",
        "# # the .flow() command below generates batches of randomly transformed images\n",
        "# # and saves the results to the `preview/` directory\n",
        "# path_aug = '/content/aug_img'\n",
        "# i = 0\n",
        "# for batch in datagen.flow(x, batch_size=1,\n",
        "#                           save_to_dir=path_aug, save_prefix='vertikal', save_format='jpeg'):\n",
        "#     i += 1\n",
        "#     if i > 1:\n",
        "#         break  # otherwise the generator would loop indefinitely"
      ],
      "metadata": {
        "id": "03fYOhUuwKLS"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# create model architecture"
      ],
      "metadata": {
        "id": "GRETyJMHirX4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import tensorflow as tf\n",
        "import keras_preprocessing\n",
        "from keras_preprocessing import image\n",
        "from keras_preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.models import Sequential\n",
        "from keras.layers import Conv2D, Dense, Flatten, MaxPooling2D, Dropout\n",
        "from tensorflow.keras import activations\n",
        "\n",
        "# model = Sequential()\n",
        "\n",
        "# # Layer CNN pertama\n",
        "# model.add(Conv2D(filters=32, kernel_size=3, activation='relu', input_shape=(150,150,3)))\n",
        "# model.add(MaxPool2D(pool_size=(2, 2)))\n",
        "# model.add(Dropout(rate=0.25))\n",
        "\n",
        "# # Layer CNN kedua\n",
        "# model.add(Conv2D(filters=64, kernel_size=3, activation='relu'))\n",
        "# model.add(MaxPool2D(pool_size=(2, 2)))\n",
        "# model.add(Dropout(rate=0.25))\n",
        "\n",
        "# # Layer CNN ketiga\n",
        "# model.add(Conv2D(filters=64, kernel_size=3, activation='relu'))\n",
        "\n",
        "# model.summary()"
      ],
      "metadata": {
        "id": "1k1tFggfhiUs"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# model = tf.keras.models.Sequential([\n",
        "#     tf.keras.layers.Conv2D(32, (3,3), activation='relu', input_shape=(256, 256, 3)),\n",
        "#     tf.keras.layers.MaxPooling2D(2,2),\n",
        "#     tf.keras.layers.Flatten(),\n",
        "#     tf.keras.layers.Dense(128, activation='relu'),\n",
        "#     tf.keras.layers.Dropout(0.2),\n",
        "#     tf.keras.layers.Dense(3, activation='softmax')\n",
        "# ])\n",
        "\n",
        "# model.compile(optimizer='adam',\n",
        "#               loss='categorical_crossentropy',\n",
        "#               metrics=['accuracy'])\n",
        "\n",
        "# model.summary()"
      ],
      "metadata": {
        "id": "vKbfclRii0z5"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# from tensorflow import keras\n",
        "# from tensorflow.keras import layers, models\n",
        "# from tensorflow.keras.models import Sequential\n",
        "# from keras.layers import Conv2D, Dense, Flatten, MaxPool2D, Dropout, BatchNormalization\n",
        "\n",
        "# model = keras.Sequential()\n",
        "# model.add(layers.Conv2D(filters=16, kernel_size=(3, 3), padding='same', activation = 'relu', input_shape = (256, 256, 3), kernel_initializer=\"he_normal\"))\n",
        "# model.add(layers.MaxPooling2D())\n",
        "# model.add(layers.Conv2D(filters=32, kernel_size=(3, 3), padding='same', activation = 'relu', input_shape = (256, 256, 3), kernel_initializer=\"he_normal\"))\n",
        "# model.add(layers.MaxPooling2D())\n",
        "\n",
        "# model.add(Dropout(0.25))\n",
        "# model.add(layers.Flatten())\n",
        "# model.add(BatchNormalization())\n",
        "# model.add(layers.Dense(units = 128, activation = 'relu', kernel_initializer=\"he_normal\"))\n",
        "# model.add(BatchNormalization())\n",
        "# model.add(layers.Dense(units = 64, activation = 'relu'))\n",
        "# model.add(layers.Dense(units = 3, activation = 'softmax'))\n",
        "\n",
        "# model.compile(optimizer='adam', loss='categorical_crossentropy', metrics = ['accuracy'])\n",
        "\n",
        "# model.summary()"
      ],
      "metadata": {
        "id": "VViy9RIdRSQ9"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def build_model():\n",
        "    \n",
        "    '''Sequential Model creation'''\n",
        "    model = Sequential()\n",
        "    \n",
        "    model.add(Conv2D(16,(3,3),padding='same',input_shape = (256,256,3),activation='relu'))\n",
        "    \n",
        "    model.add(Conv2D(32,(3,3),padding='same',activation='relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2,2),strides=2,padding = 'same'))\n",
        "    \n",
        "    model.add(Conv2D(32,(3,3),padding='same',activation='relu'))\n",
        "    \n",
        "    \n",
        "    model.add(Conv2D(64,(3,3),padding='same',activation='relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2,2),strides=2,padding = 'same'))\n",
        "    \n",
        "    model.add(Conv2D(64,(3,3),padding='same',activation='relu'))\n",
        "    \n",
        "    model.add(Conv2D(128,(3,3),padding='same',activation='relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2,2),strides=2,padding = 'same'))\n",
        "    \n",
        "    model.add(Conv2D(128,(3,3),padding='same',activation='relu'))\n",
        "    \n",
        "    model.add(Conv2D(128,(3,3),padding='same',activation='relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2,2),strides=2,padding = 'same'))\n",
        "    \n",
        "    \n",
        "    model.add(Flatten())\n",
        "    model.add(Dense(32))\n",
        "    model.add(Dense(32))\n",
        "    model.add(Dense(32))\n",
        "    model.add(Dense(32))\n",
        "    model.add(Dense(units = 3, activation = 'softmax'))\n",
        "    \n",
        "    return model\n",
        "\n",
        "model = build_model()\n",
        "model.summary()"
      ],
      "metadata": {
        "id": "LBo0yzua310I",
        "outputId": "986b7b53-b0b7-4d54-de65-a29628657884",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d (Conv2D)             (None, 256, 256, 16)      448       \n",
            "                                                                 \n",
            " conv2d_1 (Conv2D)           (None, 256, 256, 32)      4640      \n",
            "                                                                 \n",
            " max_pooling2d (MaxPooling2D  (None, 128, 128, 32)     0         \n",
            " )                                                               \n",
            "                                                                 \n",
            " conv2d_2 (Conv2D)           (None, 128, 128, 32)      9248      \n",
            "                                                                 \n",
            " conv2d_3 (Conv2D)           (None, 128, 128, 64)      18496     \n",
            "                                                                 \n",
            " max_pooling2d_1 (MaxPooling  (None, 64, 64, 64)       0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " conv2d_4 (Conv2D)           (None, 64, 64, 64)        36928     \n",
            "                                                                 \n",
            " conv2d_5 (Conv2D)           (None, 64, 64, 128)       73856     \n",
            "                                                                 \n",
            " max_pooling2d_2 (MaxPooling  (None, 32, 32, 128)      0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " conv2d_6 (Conv2D)           (None, 32, 32, 128)       147584    \n",
            "                                                                 \n",
            " conv2d_7 (Conv2D)           (None, 32, 32, 128)       147584    \n",
            "                                                                 \n",
            " max_pooling2d_3 (MaxPooling  (None, 16, 16, 128)      0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 32768)             0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 32)                1048608   \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 32)                1056      \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 32)                1056      \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 32)                1056      \n",
            "                                                                 \n",
            " dense_4 (Dense)             (None, 3)                 99        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 1,490,659\n",
            "Trainable params: 1,490,659\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# modelling"
      ],
      "metadata": {
        "id": "Sq_GNC6LrEie"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# train_len = 0\n",
        "# for foldername in os.listdir('/content/alzheimer-mri-dataset/Alzheimers-ADNI/train'):\n",
        "#     train_len = train_len + len(os.listdir(os.path.join('/content/alzheimer-mri-dataset/Alzheimers-ADNI/train',foldername)))\n",
        "    \n",
        "# val_len = 0\n",
        "# for foldername in os.listdir('/content/alzheimer-mri-dataset/Alzheimers-ADNI/test'):\n",
        "#     val_len = val_len + len(os.listdir(os.path.join('/content/alzheimer-mri-dataset/Alzheimers-ADNI/test',foldername)))\n",
        "\n",
        "# print(train_len)\n",
        "# print(val_len)"
      ],
      "metadata": {
        "id": "Ub28diD3rKv8"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.wrappers.scikit_learn import KerasClassifier\n",
        "import keras.backend as K"
      ],
      "metadata": {
        "id": "BVz3SZ8dnCpo"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def f1_score(y_true, y_pred): #taken from old keras source code\n",
        "    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
        "    possible_positives = K.sum(K.round(K.clip(y_true, 0, 1)))\n",
        "    predicted_positives = K.sum(K.round(K.clip(y_pred, 0, 1)))\n",
        "    precision = true_positives / (predicted_positives + K.epsilon())\n",
        "    recall = true_positives / (possible_positives + K.epsilon())\n",
        "    f1_val = 2*(precision*recall)/(precision+recall+K.epsilon())\n",
        "    return f1_val"
      ],
      "metadata": {
        "id": "fdwmMU3i3u45"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "METRICS = [\n",
        "      tf.keras.metrics.BinaryAccuracy(name='accuracy'),\n",
        "      tf.keras.metrics.Precision(name='precision'),\n",
        "      tf.keras.metrics.Recall(name='recall'),  \n",
        "      tf.keras.metrics.AUC(name='auc'),\n",
        "        f1_score,\n",
        "]"
      ],
      "metadata": {
        "id": "yRwHEivD4CF6"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics = METRICS)"
      ],
      "metadata": {
        "id": "hBC6fKzSmodM"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_history = model.fit(train_generator,\n",
        "                        validation_data=val_generator,\n",
        "                        epochs = 30,\n",
        "                        verbose = 1)"
      ],
      "metadata": {
        "id": "3uAsHW-5rDZ8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "99d9a0d1-3f51-4c2e-9c33-0d0af1fdae67"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/30\n",
            "16/16 [==============================] - 85s 5s/step - loss: 1.2157 - accuracy: 0.6586 - precision: 0.3125 - recall: 0.0202 - auc: 0.5164 - f1_score: 0.0195 - val_loss: 1.1236 - val_accuracy: 0.6667 - val_precision: 0.0000e+00 - val_recall: 0.0000e+00 - val_auc: 0.3988 - val_f1_score: 0.0000e+00\n",
            "Epoch 2/30\n",
            "16/16 [==============================] - 72s 5s/step - loss: 1.0002 - accuracy: 0.6962 - precision: 0.6897 - recall: 0.1613 - auc: 0.6899 - f1_score: 0.2086 - val_loss: 1.9381 - val_accuracy: 0.5628 - val_precision: 0.2889 - val_recall: 0.2131 - val_auc: 0.3802 - val_f1_score: 0.2333\n",
            "Epoch 3/30\n",
            "16/16 [==============================] - 72s 4s/step - loss: 0.8751 - accuracy: 0.7406 - precision: 0.6687 - recall: 0.4395 - auc: 0.7829 - f1_score: 0.5140 - val_loss: 3.1560 - val_accuracy: 0.5027 - val_precision: 0.1739 - val_recall: 0.1311 - val_auc: 0.4024 - val_f1_score: 0.1464\n",
            "Epoch 4/30\n",
            "16/16 [==============================] - 72s 5s/step - loss: 0.7855 - accuracy: 0.7621 - precision: 0.7101 - recall: 0.4839 - auc: 0.8250 - f1_score: 0.5707 - val_loss: 5.3948 - val_accuracy: 0.4699 - val_precision: 0.1897 - val_recall: 0.1803 - val_auc: 0.3835 - val_f1_score: 0.1875\n",
            "Epoch 5/30\n",
            "16/16 [==============================] - 72s 5s/step - loss: 0.7433 - accuracy: 0.7769 - precision: 0.7356 - recall: 0.5161 - auc: 0.8602 - f1_score: 0.5744 - val_loss: 1.3105 - val_accuracy: 0.6448 - val_precision: 0.3750 - val_recall: 0.0984 - val_auc: 0.3518 - val_f1_score: 0.1412\n",
            "Epoch 6/30\n",
            "16/16 [==============================] - 75s 5s/step - loss: 0.7645 - accuracy: 0.8078 - precision: 0.8107 - recall: 0.5524 - auc: 0.8797 - f1_score: 0.6173 - val_loss: 1.4860 - val_accuracy: 0.5574 - val_precision: 0.1875 - val_recall: 0.0984 - val_auc: 0.3749 - val_f1_score: 0.1271\n",
            "Epoch 7/30\n",
            "14/16 [=========================>....] - ETA: 8s - loss: 0.6725 - accuracy: 0.8363 - precision: 0.8314 - recall: 0.6384 - auc: 0.8981 - f1_score: 0.7156 "
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Summarize history for loss\n",
        "\n",
        "plt.plot(model_history.history['loss'])\n",
        "plt.plot(model_history.history['val_loss'])\n",
        "plt.title('Model Loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Validation'], loc='upper left', bbox_to_anchor=(1,1))\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "XS0yGh1gL-SR",
        "outputId": "2d242f88-fff5-4808-c3e2-686a200da1e9"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Summarize history for loss\n",
        "\n",
        "plt.plot(model_history.history['accuracy'])\n",
        "plt.plot(model_history.history['val_accuracy'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Validation'], loc='upper left', bbox_to_anchor=(1,1))\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "0Z2kzTsxY--e",
        "outputId": "2cc90301-5daf-4b64-d280-36eaf7aa1143"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#%% PLOTTING RESULTS (Train vs Validation FOLDER 1)\n",
        "\n",
        "def Train_Val_Plot(acc,val_acc,loss,val_loss,auc,val_auc,precision,val_precision,f1,val_f1):\n",
        "    \n",
        "    fig, (ax1, ax2,ax3,ax4,ax5) = plt.subplots(1,5, figsize= (20,5))\n",
        "    fig.suptitle(\" MODEL'S METRICS VISUALIZATION \")\n",
        "\n",
        "    ax1.plot(range(1, len(acc) + 1), acc)\n",
        "    ax1.plot(range(1, len(val_acc) + 1), val_acc)\n",
        "    ax1.set_title('History of Accuracy')\n",
        "    ax1.set_xlabel('Epochs')\n",
        "    ax1.set_ylabel('Accuracy')\n",
        "    ax1.legend(['training', 'validation'])\n",
        "\n",
        "\n",
        "    ax2.plot(range(1, len(loss) + 1), loss)\n",
        "    ax2.plot(range(1, len(val_loss) + 1), val_loss)\n",
        "    ax2.set_title('History of Loss')\n",
        "    ax2.set_xlabel('Epochs')\n",
        "    ax2.set_ylabel('Loss')\n",
        "    ax2.legend(['training', 'validation'])\n",
        "    \n",
        "    ax3.plot(range(1, len(auc) + 1), auc)\n",
        "    ax3.plot(range(1, len(val_auc) + 1), val_auc)\n",
        "    ax3.set_title('History of AUC')\n",
        "    ax3.set_xlabel('Epochs')\n",
        "    ax3.set_ylabel('AUC')\n",
        "    ax3.legend(['training', 'validation'])\n",
        "    \n",
        "    ax4.plot(range(1, len(precision) + 1), precision)\n",
        "    ax4.plot(range(1, len(val_precision) + 1), val_precision)\n",
        "    ax4.set_title('History of Precision')\n",
        "    ax4.set_xlabel('Epochs')\n",
        "    ax4.set_ylabel('Precision')\n",
        "    ax4.legend(['training', 'validation'])\n",
        "    \n",
        "    ax5.plot(range(1, len(f1) + 1), f1)\n",
        "    ax5.plot(range(1, len(val_f1) + 1), val_f1)\n",
        "    ax5.set_title('History of F1-score')\n",
        "    ax5.set_xlabel('Epochs')\n",
        "    ax5.set_ylabel('F1 score')\n",
        "    ax5.legend(['training', 'validation'])\n",
        "\n",
        "\n",
        "    plt.show()\n",
        "    \n",
        "\n",
        "Train_Val_Plot(model_history.history['accuracy'],model_history.history['val_accuracy'],\n",
        "               model_history.history['loss'],model_history.history['val_loss'],\n",
        "               model_history.history['auc'],model_history.history['val_auc'],\n",
        "               model_history.history['precision'],model_history.history['val_precision'],\n",
        "               model_history.history['f1_score'],model_history.history['val_f1_score']\n",
        "              )"
      ],
      "metadata": {
        "id": "JPbHyoCN3JSk",
        "outputId": "4709c403-41b2-45fe-dbd5-5f1f1c53e49d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 348
        }
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1440x360 with 5 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction= model.predict(val_generator, steps=np.ceil(val_generator.samples/val_generator.batch_size), verbose=2)\n",
        "prediction= (prediction > 0.5)\n",
        "val_labels=val_generator.classes"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "je52vjzCZRYY",
        "outputId": "383e6193-1f38-4207-d335-e461a34cadaa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 - 1s - 971ms/epoch - 243ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import metrics\n",
        "from sklearn.metrics import confusion_matrix, accuracy_score, classification_report\n",
        "from mlxtend.plotting import plot_confusion_matrix\n",
        "\n",
        "print(classification_report(val_generator.classes, prediction))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 345
        },
        "id": "K3PVlF3VZ-Pk",
        "outputId": "a50a02bb-8138-409f-a8cd-e295cbf89d4a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "ValueError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-18-7e5551e9f6e6>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mmlxtend\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplotting\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mplot_confusion_matrix\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mclassification_report\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mval_generator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclasses\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mprediction\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36mclassification_report\u001b[0;34m(y_true, y_pred, labels, target_names, sample_weight, digits, output_dict, zero_division)\u001b[0m\n\u001b[1;32m   2108\u001b[0m     \"\"\"\n\u001b[1;32m   2109\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2110\u001b[0;31m     \u001b[0my_type\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_check_targets\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2111\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2112\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mlabels\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36m_check_targets\u001b[0;34m(y_true, y_pred)\u001b[0m\n\u001b[1;32m     93\u001b[0m         raise ValueError(\n\u001b[1;32m     94\u001b[0m             \"Classification metrics can't handle a mix of {0} and {1} targets\".format(\n\u001b[0;32m---> 95\u001b[0;31m                 \u001b[0mtype_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtype_pred\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     96\u001b[0m             )\n\u001b[1;32m     97\u001b[0m         )\n",
            "\u001b[0;31mValueError\u001b[0m: Classification metrics can't handle a mix of multiclass and multilabel-indicator targets"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "VOEIdpD7aBMB"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}