Plant_Leaf_Diseases_Classification_with_Quantized__TFLite_Model.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Plant Leaf Diseases Classification.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyMDezx0E1Wwm7KYw6krX98l",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Vidi005/IDCamp-Dicoding-Challenge-SDGs-2020/blob/master/Plant_Leaf_Diseases_Classification_with_Quantized__TFLite_Model.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fJEscDjOtgYH",
        "colab_type": "text"
      },
      "source": [
        "# **Farm Doctor**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KcHddQ17tyef",
        "colab_type": "text"
      },
      "source": [
        "Program ML untuk klasifikasi penyakit pada daun tanaman\n",
        "<br>\n",
        "Program ini dibuat menggunakan CNN dengan library Keras dan backend TensorFlow."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GFyhoUc4uD88",
        "colab_type": "text"
      },
      "source": [
        "## Memasang Kaggle untuk Mengimpor Dataset dari Kaggle"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hj8zX4F02RhK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        },
        "outputId": "e6a5fe47-48d5-472d-fd21-532ae03da9a8"
      },
      "source": [
        "!pip install kaggle"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: kaggle in /usr/local/lib/python3.6/dist-packages (1.5.6)\n",
            "Requirement already satisfied: six>=1.10 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.15.0)\n",
            "Requirement already satisfied: python-slugify in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.0.1)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.23.0)\n",
            "Requirement already satisfied: python-dateutil in /usr/local/lib/python3.6/dist-packages (from kaggle) (2.8.1)\n",
            "Requirement already satisfied: urllib3<1.25,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from kaggle) (1.24.3)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from kaggle) (4.41.1)\n",
            "Requirement already satisfied: certifi in /usr/local/lib/python3.6/dist-packages (from kaggle) (2020.6.20)\n",
            "Requirement already satisfied: text-unidecode>=1.3 in /usr/local/lib/python3.6/dist-packages (from python-slugify->kaggle) (1.3)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (3.0.4)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->kaggle) (2.10)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XAoIQrUnvlur",
        "colab_type": "text"
      },
      "source": [
        "## Upload Kaggle API"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CXaTxXNu3W8_",
        "colab_type": "code",
        "colab": {
          "resources": {
            "http:https://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 89
        },
        "outputId": "e7a1864e-c0b6-46b3-bbea-9d5452852fe8"
      },
      "source": [
        "from google.colab import files\n",
        "files.upload()"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-8da66b59-ddbb-4aa8-97ea-74721796dbee\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-8da66b59-ddbb-4aa8-97ea-74721796dbee\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving kaggle.json to kaggle.json\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'kaggle.json': b'{\"username\":\"vidi005\",\"key\":\"9fce1c2f299c8e658ea0b78f1cce224f\"}'}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fNzNljrAvrCt",
        "colab_type": "text"
      },
      "source": [
        "## Menyimpan File Kaggle API untuk Download Datasets"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FbXfW2P63oMC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!mkdir -p ~/.kaggle\n",
        "!cp kaggle.json ~/.kaggle/\n",
        "!chmod 600 ~/.kaggle/kaggle.json"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pkjtuId6xpJB",
        "colab_type": "text"
      },
      "source": [
        "## Preprocessing Datasets"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C_RLcltWv86V",
        "colab_type": "text"
      },
      "source": [
        "### Mendwonload Dataset Plant Diseases"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SSjTIHNx0uzV",
        "colab_type": "text"
      },
      "source": [
        "Dataset didownload dari link berikut:\n",
        "<br>\n",
        "https://www.kaggle.com/daoliu/plant-622"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NnFtBLlP3438",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "5f498448-b3d9-4c52-c20e-a2e12bf1e18e"
      },
      "source": [
        "!kaggle datasets download -d daoliu/plant-622"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading plant-622.zip to /content\n",
            " 99% 1.32G/1.34G [00:11<00:00, 122MB/s]\n",
            "100% 1.34G/1.34G [00:11<00:00, 128MB/s]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jd2JMAlZwIbP",
        "colab_type": "text"
      },
      "source": [
        "### Menampilkan List Folder Content"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "moR0EgVk5GyW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "16a2ad1c-b91a-45f4-8808-450080c9c39e"
      },
      "source": [
        "!ls"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "kaggle.json  plant-622.zip  sample_data\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Hxcu2NQ4wSl6",
        "colab_type": "text"
      },
      "source": [
        "### Mengekstrak Dataset yang Sudah didownload"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HpJ3OZ3l5ad2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Menyimpan hasil ekstrak ke dalam folder datasets\n",
        "import zipfile,os\n",
        "local_zip = \"plant-622.zip\"\n",
        "zip_ref = zipfile.ZipFile(local_zip, \"r\")\n",
        "zip_ref.extractall(\"/content/datasets\")\n",
        "zip_ref.close()"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G39aprg3SKYB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Menghapus folder dataset pelatihan yang tidak diperlukan\n",
        "import shutil\n",
        "shutil.rmtree(\"/content/datasets/train/Blueberry___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/train/Orange___Haunglongbing_(Citrus_greening)\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/train/Peach___Bacterial_spot\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/train/Peach___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/train/Raspberry___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/train/Soybean___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/train/Squash___Powdery_mildew\", ignore_errors=False, onerror=None)\n",
        "\n",
        "#Menghapus dataset validasi yang tidak diperlukan\n",
        "#Dataset validasi yang digunakan diambil 20% dari dataset pelatihan menggunakan image generator\n",
        "shutil.rmtree(\"/content/datasets/valid\", ignore_errors=False, onerror=None)\n",
        "\n",
        "#Menghapus dataset pengujian yang tidak diperlukan\n",
        "shutil.rmtree(\"/content/datasets/test/Blueberry___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/test/Orange___Haunglongbing_(Citrus_greening)\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/test/Peach___Bacterial_spot\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/test/Peach___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/test/Raspberry___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/test/Soybean___healthy\", ignore_errors=False, onerror=None)\n",
        "shutil.rmtree(\"/content/datasets/test/Squash___Powdery_mildew\", ignore_errors=False, onerror=None)"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ML9_IgtoxI4V",
        "colab_type": "text"
      },
      "source": [
        "## Mengimpor Libraries yang dibutuhkan"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YsFrRLdQ8QHA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 190
        },
        "outputId": "9bc49bfa-2fd6-4613-8740-e855c1a5a314"
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from sklearn.model_selection import train_test_split\n",
        "from skimage.io import imread\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras.callbacks import EarlyStopping\n",
        "import numpy as np\n",
        "from skimage.transform import resize\n",
        "from keras.utils.np_utils import to_categorical\n",
        "\n",
        "# Mengihitung waktu lamanya eksekusi tiap sel di Google Colab\n",
        "!pip install ipython-autotime\n",
        "\n",
        "%load_ext autotime"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting ipython-autotime\n",
            "  Downloading https://files.pythonhosted.org/packages/e6/f9/0626bbdb322e3a078d968e87e3b01341e7890544de891d0cb613641220e6/ipython-autotime-0.1.tar.bz2\n",
            "Building wheels for collected packages: ipython-autotime\n",
            "  Building wheel for ipython-autotime (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for ipython-autotime: filename=ipython_autotime-0.1-cp36-none-any.whl size=1831 sha256=2891e14266f4e91d28d79817c7c9306e083d2c525761a25d92da4b75c5cdfec1\n",
            "  Stored in directory: /root/.cache/pip/wheels/d2/df/81/2db1e54bc91002cec40334629bc39cfa86dff540b304ebcd6e\n",
            "Successfully built ipython-autotime\n",
            "Installing collected packages: ipython-autotime\n",
            "Successfully installed ipython-autotime-0.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_kFAzksXxVGu",
        "colab_type": "text"
      },
      "source": [
        "## Menampilkan Versi TensorFlow yang digunakan"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k8MNFKcb8WBZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "29ea40c2-ccc2-4553-ee9f-0fc2a33ae1bb"
      },
      "source": [
        "print(tf.__version__)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.3.0\n",
            "time: 1.53 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i52zugYyxd2h",
        "colab_type": "text"
      },
      "source": [
        "## Preprocessing Datasets"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KI8DeSFEx2G5",
        "colab_type": "text"
      },
      "source": [
        "### Membuat Objek ImageDataGenerator & Data Augmentation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cU_ZuM4vyHdI",
        "colab_type": "text"
      },
      "source": [
        "Membuat obyek gambar dataset dengan ImageDataGenerator, Autmentasi dataset, dan Membagi dataset validasi (validation) sebesar 20% dan dataset pelatihan (training) sebesar 80%."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QLR5yo1L7z6J",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f2565f69-f2e5-4d97-dca4-e76ace851770"
      },
      "source": [
        "datagen = tf.keras.preprocessing.image.ImageDataGenerator(\n",
        "    rescale=1./255, \n",
        "    validation_split = 0.2,\n",
        "    rotation_range = 20,\n",
        "    horizontal_flip = True,\n",
        "    shear_range = 0.2,\n",
        "    zoom_range = 0.2,    \n",
        "    vertical_flip = True,\n",
        "    fill_mode = \"nearest\")"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "time: 2.24 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fPtbXeyYyMHj",
        "colab_type": "text"
      },
      "source": [
        "### Menggunakan Obyek ImageDataGenerator"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "spCgvam78OSs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        },
        "outputId": "1757abbf-b975-4bd8-b401-a3885baa7c8e"
      },
      "source": [
        "#Memuat dataset pelatihan & validasi\n",
        "base_dir = os.path.join(\"datasets/train\")\n",
        "\n",
        "img_size = 224\n",
        "batch_size = 128\n",
        "\n",
        "train_generator = datagen.flow_from_directory(\n",
        "    base_dir,\n",
        "    target_size=(img_size, img_size),\n",
        "    batch_size=batch_size, \n",
        "    subset='training',\n",
        "    class_mode= 'categorical')\n",
        "\n",
        "val_generator = datagen.flow_from_directory(\n",
        "    base_dir,\n",
        "    target_size=(img_size, img_size),\n",
        "    batch_size=batch_size, \n",
        "    subset='validation',\n",
        "    class_mode= 'categorical')\n",
        "\n",
        "#Memuat dataset pengujian\n",
        "X_test = []\n",
        "y_test = []\n",
        "labels = ['Apple___Apple_scab',\n",
        "          'Apple___Black_rot',\n",
        "          'Apple___Cedar_apple_rust',\n",
        "          'Apple___healthy',\n",
        "          'Cherry_(including_sour)___Powdery_mildew',\n",
        "          'Cherry_(including_sour)___healthy',\n",
        "          'Corn_(maize)___Cercospora_leaf_spot Gray_leaf_spot',\n",
        "          'Corn_(maize)___Common_rust_',\n",
        "          'Corn_(maize)___Northern_Leaf_Blight',\n",
        "          'Corn_(maize)___healthy',\n",
        "          'Grape___Black_rot',\n",
        "          'Grape___Esca_(Black_Measles)',\n",
        "          'Grape___Leaf_blight_(Isariopsis_Leaf_Spot)',\n",
        "          'Grape___healthy',\n",
        "          'Pepper,_bell___Bacterial_spot',\n",
        "          'Pepper,_bell___healthy',\n",
        "          'Potato___Early_blight',\n",
        "          'Potato___Late_blight',\n",
        "          'Potato___healthy',\n",
        "          'Strawberry___Leaf_scorch',\n",
        "          'Strawberry___Leaf_scorch',\n",
        "          'Tomato___Bacterial_spot',\n",
        "          'Tomato___Early_blight',\n",
        "          'Tomato___Late_blight',\n",
        "          'Tomato___Leaf_Mold',\n",
        "          'Tomato___Septoria_leaf_spot',\n",
        "          'Tomato___Spider_mites Two-spotted_spider_mite',\n",
        "          'Tomato___Target_Spot',\n",
        "          'Tomato___Tomato_Yellow_Leaf_Curl_Virus',\n",
        "          'Tomato___Tomato_mosaic_virus',\n",
        "          'Tomato___healthy']\n",
        "\n",
        "for i,label in enumerate(labels):\n",
        "    folder = os.path.join(\"datasets/test\",label)\n",
        "    files = sorted(os.listdir(folder))\n",
        "    files = [x for x in files if x.endswith(\".jpg\")]\n",
        "    for k,file in enumerate(files):\n",
        "        image_path = os.path.join(folder, file)\n",
        "        \n",
        "        image = imread(image_path)/255.\n",
        "        image = resize(image,(224,224))\n",
        "        X_test.append(image)\n",
        "        category = os.path.split(folder)[-1]\n",
        "        y_test.append(i)\n",
        "\n",
        "X_test = np.array(X_test)\n",
        "y_test = np.array(y_test)\n",
        "\n",
        "#Menampilkan bentuk dari masing-masing dataset\n",
        "for image_batch, label_batch in train_generator:\n",
        "  break\n",
        "print(\"Bentuk array dari dataset train (pelatihan) adalah:\", image_batch.shape,label_batch.shape)\n",
        "for image_batch, label_batch in val_generator:\n",
        "  break\n",
        "print(\"Bentuk array dari dataset validation (validasi) adalah:\", image_batch.shape,label_batch.shape)\n",
        "print(\"Bentuk array dari dataset test (pengujian) adalah:\", X_test.shape,y_test.shape)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 34429 images belonging to 31 classes.\n",
            "Found 8595 images belonging to 31 classes.\n",
            "Bentuk array dari dataset train (pelatihan) adalah: (128, 224, 224, 3) (128, 31)\n",
            "Bentuk array dari dataset validation (validasi) adalah: (128, 224, 224, 3) (128, 31)\n",
            "Bentuk array dari dataset test (pengujian) adalah: (601, 224, 224, 3) (601,)\n",
            "time: 15 s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cIdGru14yhrJ",
        "colab_type": "text"
      },
      "source": [
        "## Menyimpan Label Datasets"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pDcFgxYu895P",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 71
        },
        "outputId": "e386e4b5-bec7-4391-ee13-7887b8ea4223"
      },
      "source": [
        "print (train_generator.class_indices)\n",
        "\n",
        "labels_txt = '\\n'.join(sorted(train_generator.class_indices.keys()))\n",
        "\n",
        "with open('labels.txt', 'w') as f:\n",
        "  f.write(labels_txt)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{'Apple___Apple_scab': 0, 'Apple___Black_rot': 1, 'Apple___Cedar_apple_rust': 2, 'Apple___healthy': 3, 'Cherry_(including_sour)___Powdery_mildew': 4, 'Cherry_(including_sour)___healthy': 5, 'Corn_(maize)___Cercospora_leaf_spot Gray_leaf_spot': 6, 'Corn_(maize)___Common_rust_': 7, 'Corn_(maize)___Northern_Leaf_Blight': 8, 'Corn_(maize)___healthy': 9, 'Grape___Black_rot': 10, 'Grape___Esca_(Black_Measles)': 11, 'Grape___Leaf_blight_(Isariopsis_Leaf_Spot)': 12, 'Grape___healthy': 13, 'Pepper,_bell___Bacterial_spot': 14, 'Pepper,_bell___healthy': 15, 'Potato___Early_blight': 16, 'Potato___Late_blight': 17, 'Potato___healthy': 18, 'Strawberry___Leaf_scorch': 19, 'Strawberry___healthy': 20, 'Tomato___Bacterial_spot': 21, 'Tomato___Early_blight': 22, 'Tomato___Late_blight': 23, 'Tomato___Leaf_Mold': 24, 'Tomato___Septoria_leaf_spot': 25, 'Tomato___Spider_mites Two-spotted_spider_mite': 26, 'Tomato___Target_Spot': 27, 'Tomato___Tomato_Yellow_Leaf_Curl_Virus': 28, 'Tomato___Tomato_mosaic_virus': 29, 'Tomato___healthy': 30}\n",
            "time: 3.17 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mKW3wxvGytma",
        "colab_type": "text"
      },
      "source": [
        "## Membuat Model dari Jaringan CNN yang Sudah dipelajari Sebelumnya (pre-trained convnets)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BsNzogMa9ZQx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        },
        "outputId": "3cfcce20-0550-4a09-f097-e5826634a80b"
      },
      "source": [
        "IMG_SHAPE = (224, 224, 3)\n",
        "# Membuat model dasar (base model) dari pre-trained model MobileNet\n",
        "base_model = tf.keras.applications.MobileNet(input_shape=IMG_SHAPE,\n",
        "                                              include_top=False, \n",
        "                                              weights='imagenet')"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet/mobilenet_1_0_224_tf_no_top.h5\n",
            "17227776/17225924 [==============================] - 0s 0us/step\n",
            "time: 6.71 s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nU4xTUdFy0FS",
        "colab_type": "text"
      },
      "source": [
        "### Feature Extraction"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wDlRKJ6U9k9C",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "0a1e6f9c-0b32-4990-e6c1-d37bdc6e34a3"
      },
      "source": [
        "base_model.trainable = False\n",
        "base_model.summary()"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"mobilenet_1.00_224\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_1 (InputLayer)         [(None, 224, 224, 3)]     0         \n",
            "_________________________________________________________________\n",
            "conv1_pad (ZeroPadding2D)    (None, 225, 225, 3)       0         \n",
            "_________________________________________________________________\n",
            "conv1 (Conv2D)               (None, 112, 112, 32)      864       \n",
            "_________________________________________________________________\n",
            "conv1_bn (BatchNormalization (None, 112, 112, 32)      128       \n",
            "_________________________________________________________________\n",
            "conv1_relu (ReLU)            (None, 112, 112, 32)      0         \n",
            "_________________________________________________________________\n",
            "conv_dw_1 (DepthwiseConv2D)  (None, 112, 112, 32)      288       \n",
            "_________________________________________________________________\n",
            "conv_dw_1_bn (BatchNormaliza (None, 112, 112, 32)      128       \n",
            "_________________________________________________________________\n",
            "conv_dw_1_relu (ReLU)        (None, 112, 112, 32)      0         \n",
            "_________________________________________________________________\n",
            "conv_pw_1 (Conv2D)           (None, 112, 112, 64)      2048      \n",
            "_________________________________________________________________\n",
            "conv_pw_1_bn (BatchNormaliza (None, 112, 112, 64)      256       \n",
            "_________________________________________________________________\n",
            "conv_pw_1_relu (ReLU)        (None, 112, 112, 64)      0         \n",
            "_________________________________________________________________\n",
            "conv_pad_2 (ZeroPadding2D)   (None, 113, 113, 64)      0         \n",
            "_________________________________________________________________\n",
            "conv_dw_2 (DepthwiseConv2D)  (None, 56, 56, 64)        576       \n",
            "_________________________________________________________________\n",
            "conv_dw_2_bn (BatchNormaliza (None, 56, 56, 64)        256       \n",
            "_________________________________________________________________\n",
            "conv_dw_2_relu (ReLU)        (None, 56, 56, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv_pw_2 (Conv2D)           (None, 56, 56, 128)       8192      \n",
            "_________________________________________________________________\n",
            "conv_pw_2_bn (BatchNormaliza (None, 56, 56, 128)       512       \n",
            "_________________________________________________________________\n",
            "conv_pw_2_relu (ReLU)        (None, 56, 56, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_3 (DepthwiseConv2D)  (None, 56, 56, 128)       1152      \n",
            "_________________________________________________________________\n",
            "conv_dw_3_bn (BatchNormaliza (None, 56, 56, 128)       512       \n",
            "_________________________________________________________________\n",
            "conv_dw_3_relu (ReLU)        (None, 56, 56, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_3 (Conv2D)           (None, 56, 56, 128)       16384     \n",
            "_________________________________________________________________\n",
            "conv_pw_3_bn (BatchNormaliza (None, 56, 56, 128)       512       \n",
            "_________________________________________________________________\n",
            "conv_pw_3_relu (ReLU)        (None, 56, 56, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv_pad_4 (ZeroPadding2D)   (None, 57, 57, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_4 (DepthwiseConv2D)  (None, 28, 28, 128)       1152      \n",
            "_________________________________________________________________\n",
            "conv_dw_4_bn (BatchNormaliza (None, 28, 28, 128)       512       \n",
            "_________________________________________________________________\n",
            "conv_dw_4_relu (ReLU)        (None, 28, 28, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_4 (Conv2D)           (None, 28, 28, 256)       32768     \n",
            "_________________________________________________________________\n",
            "conv_pw_4_bn (BatchNormaliza (None, 28, 28, 256)       1024      \n",
            "_________________________________________________________________\n",
            "conv_pw_4_relu (ReLU)        (None, 28, 28, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_5 (DepthwiseConv2D)  (None, 28, 28, 256)       2304      \n",
            "_________________________________________________________________\n",
            "conv_dw_5_bn (BatchNormaliza (None, 28, 28, 256)       1024      \n",
            "_________________________________________________________________\n",
            "conv_dw_5_relu (ReLU)        (None, 28, 28, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_5 (Conv2D)           (None, 28, 28, 256)       65536     \n",
            "_________________________________________________________________\n",
            "conv_pw_5_bn (BatchNormaliza (None, 28, 28, 256)       1024      \n",
            "_________________________________________________________________\n",
            "conv_pw_5_relu (ReLU)        (None, 28, 28, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv_pad_6 (ZeroPadding2D)   (None, 29, 29, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_6 (DepthwiseConv2D)  (None, 14, 14, 256)       2304      \n",
            "_________________________________________________________________\n",
            "conv_dw_6_bn (BatchNormaliza (None, 14, 14, 256)       1024      \n",
            "_________________________________________________________________\n",
            "conv_dw_6_relu (ReLU)        (None, 14, 14, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_6 (Conv2D)           (None, 14, 14, 512)       131072    \n",
            "_________________________________________________________________\n",
            "conv_pw_6_bn (BatchNormaliza (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_pw_6_relu (ReLU)        (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_7 (DepthwiseConv2D)  (None, 14, 14, 512)       4608      \n",
            "_________________________________________________________________\n",
            "conv_dw_7_bn (BatchNormaliza (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_dw_7_relu (ReLU)        (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_7 (Conv2D)           (None, 14, 14, 512)       262144    \n",
            "_________________________________________________________________\n",
            "conv_pw_7_bn (BatchNormaliza (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_pw_7_relu (ReLU)        (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_8 (DepthwiseConv2D)  (None, 14, 14, 512)       4608      \n",
            "_________________________________________________________________\n",
            "conv_dw_8_bn (BatchNormaliza (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_dw_8_relu (ReLU)        (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_8 (Conv2D)           (None, 14, 14, 512)       262144    \n",
            "_________________________________________________________________\n",
            "conv_pw_8_bn (BatchNormaliza (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_pw_8_relu (ReLU)        (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_9 (DepthwiseConv2D)  (None, 14, 14, 512)       4608      \n",
            "_________________________________________________________________\n",
            "conv_dw_9_bn (BatchNormaliza (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_dw_9_relu (ReLU)        (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_9 (Conv2D)           (None, 14, 14, 512)       262144    \n",
            "_________________________________________________________________\n",
            "conv_pw_9_bn (BatchNormaliza (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_pw_9_relu (ReLU)        (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_10 (DepthwiseConv2D) (None, 14, 14, 512)       4608      \n",
            "_________________________________________________________________\n",
            "conv_dw_10_bn (BatchNormaliz (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_dw_10_relu (ReLU)       (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_10 (Conv2D)          (None, 14, 14, 512)       262144    \n",
            "_________________________________________________________________\n",
            "conv_pw_10_bn (BatchNormaliz (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_pw_10_relu (ReLU)       (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_11 (DepthwiseConv2D) (None, 14, 14, 512)       4608      \n",
            "_________________________________________________________________\n",
            "conv_dw_11_bn (BatchNormaliz (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_dw_11_relu (ReLU)       (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_pw_11 (Conv2D)          (None, 14, 14, 512)       262144    \n",
            "_________________________________________________________________\n",
            "conv_pw_11_bn (BatchNormaliz (None, 14, 14, 512)       2048      \n",
            "_________________________________________________________________\n",
            "conv_pw_11_relu (ReLU)       (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_pad_12 (ZeroPadding2D)  (None, 15, 15, 512)       0         \n",
            "_________________________________________________________________\n",
            "conv_dw_12 (DepthwiseConv2D) (None, 7, 7, 512)         4608      \n",
            "_________________________________________________________________\n",
            "conv_dw_12_bn (BatchNormaliz (None, 7, 7, 512)         2048      \n",
            "_________________________________________________________________\n",
            "conv_dw_12_relu (ReLU)       (None, 7, 7, 512)         0         \n",
            "_________________________________________________________________\n",
            "conv_pw_12 (Conv2D)          (None, 7, 7, 1024)        524288    \n",
            "_________________________________________________________________\n",
            "conv_pw_12_bn (BatchNormaliz (None, 7, 7, 1024)        4096      \n",
            "_________________________________________________________________\n",
            "conv_pw_12_relu (ReLU)       (None, 7, 7, 1024)        0         \n",
            "_________________________________________________________________\n",
            "conv_dw_13 (DepthwiseConv2D) (None, 7, 7, 1024)        9216      \n",
            "_________________________________________________________________\n",
            "conv_dw_13_bn (BatchNormaliz (None, 7, 7, 1024)        4096      \n",
            "_________________________________________________________________\n",
            "conv_dw_13_relu (ReLU)       (None, 7, 7, 1024)        0         \n",
            "_________________________________________________________________\n",
            "conv_pw_13 (Conv2D)          (None, 7, 7, 1024)        1048576   \n",
            "_________________________________________________________________\n",
            "conv_pw_13_bn (BatchNormaliz (None, 7, 7, 1024)        4096      \n",
            "_________________________________________________________________\n",
            "conv_pw_13_relu (ReLU)       (None, 7, 7, 1024)        0         \n",
            "=================================================================\n",
            "Total params: 3,228,864\n",
            "Trainable params: 0\n",
            "Non-trainable params: 3,228,864\n",
            "_________________________________________________________________\n",
            "time: 26.8 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ex7kSTzXzE2I",
        "colab_type": "text"
      },
      "source": [
        "## Membuat Model Arsitektur CNN"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "csfkNloMzQWQ",
        "colab_type": "text"
      },
      "source": [
        "Membuat arsitektur CNN mulai dari Lapisan Masukkan (Input Layer), Lapisan Konvolusi (Convolution Layer), Lapisan aktivasi (Activation Layer), Pooling Layer, sampai Lapisan terhubung erat (Fully Connected Layer/Densely Connected Layer)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1Ji1Xs0D9or9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "78b363ce-88ec-47f2-90a1-519982ae85d0"
      },
      "source": [
        "model = tf.keras.models.Sequential([\n",
        "    base_model,    \n",
        "    tf.keras.layers.Conv2D(32, 3, activation='relu'),\n",
        "    #tf.keras.layers.Dropout(0.2),\n",
        "    tf.keras.layers.GlobalAveragePooling2D(),\n",
        "    tf.keras.layers.Dense(31, activation='softmax')\n",
        "])"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "time: 210 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fxbh9zgRzT5h",
        "colab_type": "text"
      },
      "source": [
        "### Mengkompilasi Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GdlCU1DL99RC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        },
        "outputId": "f92b6426-a32e-439f-f90b-807fb0714cbf"
      },
      "source": [
        "model.compile(\"adam\",loss=\"categorical_crossentropy\",metrics=[\"acc\"])\n",
        "model.summary()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "mobilenet_1.00_224 (Function (None, 7, 7, 1024)        3228864   \n",
            "_________________________________________________________________\n",
            "conv2d (Conv2D)              (None, 5, 5, 32)          294944    \n",
            "_________________________________________________________________\n",
            "global_average_pooling2d (Gl (None, 32)                0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 31)                1023      \n",
            "=================================================================\n",
            "Total params: 3,524,831\n",
            "Trainable params: 295,967\n",
            "Non-trainable params: 3,228,864\n",
            "_________________________________________________________________\n",
            "time: 24.8 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GRkcGbsKzbSJ",
        "colab_type": "text"
      },
      "source": [
        "### Melathi Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I_nEgtEG-KVm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 714
        },
        "outputId": "f15f5368-094d-49af-cc1d-8a231b14f980"
      },
      "source": [
        "#es = EarlyStopping(monitor=\"val_loss\", patience=7, verbose=1, min_delta=0.09, mode=\"auto\")\n",
        "history = model.fit(train_generator, \n",
        "                    epochs=20, \n",
        "                    validation_data=val_generator)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "269/269 [==============================] - 469s 2s/step - loss: 1.2467 - acc: 0.6343 - val_loss: 0.5927 - val_acc: 0.8084\n",
            "Epoch 2/20\n",
            "269/269 [==============================] - 467s 2s/step - loss: 0.4674 - acc: 0.8491 - val_loss: 0.4548 - val_acc: 0.8550\n",
            "Epoch 3/20\n",
            "269/269 [==============================] - 462s 2s/step - loss: 0.3550 - acc: 0.8845 - val_loss: 0.4003 - val_acc: 0.8707\n",
            "Epoch 4/20\n",
            "269/269 [==============================] - 462s 2s/step - loss: 0.3229 - acc: 0.8935 - val_loss: 0.3862 - val_acc: 0.8733\n",
            "Epoch 5/20\n",
            "269/269 [==============================] - 459s 2s/step - loss: 0.2971 - acc: 0.9004 - val_loss: 0.3485 - val_acc: 0.8897\n",
            "Epoch 6/20\n",
            "269/269 [==============================] - 462s 2s/step - loss: 0.2733 - acc: 0.9095 - val_loss: 0.3403 - val_acc: 0.8906\n",
            "Epoch 7/20\n",
            "269/269 [==============================] - 464s 2s/step - loss: 0.2623 - acc: 0.9129 - val_loss: 0.3247 - val_acc: 0.8961\n",
            "Epoch 8/20\n",
            "269/269 [==============================] - 459s 2s/step - loss: 0.2562 - acc: 0.9125 - val_loss: 0.3173 - val_acc: 0.8961\n",
            "Epoch 9/20\n",
            "269/269 [==============================] - 460s 2s/step - loss: 0.2394 - acc: 0.9183 - val_loss: 0.3134 - val_acc: 0.9049\n",
            "Epoch 10/20\n",
            "269/269 [==============================] - 465s 2s/step - loss: 0.2356 - acc: 0.9212 - val_loss: 0.2978 - val_acc: 0.9033\n",
            "Epoch 11/20\n",
            "269/269 [==============================] - 455s 2s/step - loss: 0.2401 - acc: 0.9192 - val_loss: 0.3278 - val_acc: 0.9011\n",
            "Epoch 12/20\n",
            "269/269 [==============================] - 454s 2s/step - loss: 0.2332 - acc: 0.9218 - val_loss: 0.3246 - val_acc: 0.9023\n",
            "Epoch 13/20\n",
            "269/269 [==============================] - 454s 2s/step - loss: 0.2209 - acc: 0.9245 - val_loss: 0.3347 - val_acc: 0.8952\n",
            "Epoch 14/20\n",
            "269/269 [==============================] - 456s 2s/step - loss: 0.2244 - acc: 0.9247 - val_loss: 0.2957 - val_acc: 0.9092\n",
            "Epoch 15/20\n",
            "269/269 [==============================] - 456s 2s/step - loss: 0.2107 - acc: 0.9283 - val_loss: 0.2819 - val_acc: 0.9106\n",
            "Epoch 16/20\n",
            "269/269 [==============================] - 455s 2s/step - loss: 0.2084 - acc: 0.9290 - val_loss: 0.3291 - val_acc: 0.9005\n",
            "Epoch 17/20\n",
            "269/269 [==============================] - 456s 2s/step - loss: 0.2026 - acc: 0.9321 - val_loss: 0.2795 - val_acc: 0.9139\n",
            "Epoch 18/20\n",
            "269/269 [==============================] - 459s 2s/step - loss: 0.2016 - acc: 0.9308 - val_loss: 0.2802 - val_acc: 0.9120\n",
            "Epoch 19/20\n",
            "269/269 [==============================] - 462s 2s/step - loss: 0.1883 - acc: 0.9363 - val_loss: 0.2618 - val_acc: 0.9173\n",
            "Epoch 20/20\n",
            "269/269 [==============================] - 473s 2s/step - loss: 0.1936 - acc: 0.9354 - val_loss: 0.2710 - val_acc: 0.9160\n",
            "time: 2h 34min 8s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0VjNo8EqzvyB",
        "colab_type": "text"
      },
      "source": [
        "### Menampilkan Grafik Model Hasil Pelatihan"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6wJ5S8ih-8_x",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 330
        },
        "outputId": "514900e2-e9cb-42c2-d413-76c5f0c0d4b7"
      },
      "source": [
        "fig = plt.figure(figsize=(7, 4))\n",
        "fig.set_figheight(10)\n",
        "fig.set_figwidth(15)\n",
        "\n",
        "plt.subplot(2, 2, 1)\n",
        "plt.plot(history.history[\"acc\"],label = \"Training Accuracy\")\n",
        "plt.plot(history.history[\"val_acc\"],label = \"Validation Accuracy\")\n",
        "plt.legend()\n",
        "plt.title(\"Kurva Tingkat Akurasi\", size=15)\n",
        "plt.grid(zorder = 0)\n",
        "\n",
        "plt.subplot(2, 2, 2)\n",
        "plt.plot(history.history[\"loss\"],label = \"Training Loss\")\n",
        "plt.plot(history.history[\"val_loss\"],label = \"Validation Loss\")\n",
        "plt.legend()\n",
        "plt.title(\"Kurva Tingkat Error\", size=15)\n",
        "plt.grid(zorder = 0)\n",
        "plt.show()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "time: 324 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dBhuf_mvz5kf",
        "colab_type": "text"
      },
      "source": [
        "## Memuat Dataset Pengujian"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GsmiaoRRR2lS",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "915f2e2c-da8a-4f20-d171-6c0294f090cf"
      },
      "source": [
        "y_test2 = to_categorical(y_test)\n",
        "X_test3, y_test3 = (X_test, y_test2)"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "time: 1.67 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IpsC9mrI0H9X",
        "colab_type": "text"
      },
      "source": [
        "## Evaluasi Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yPAc4PomSAAO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 918
        },
        "outputId": "38cbf1e6-8f3f-445b-fa5a-655fffb95853"
      },
      "source": [
        "#Menampilkan matriks yang benar dan matriks hasil prediksi\n",
        "\n",
        "#Label yang benar\n",
        "y_true = np.argmax(y_test2,axis=1)\n",
        "\n",
        "#Label prediksi\n",
        "Y_pred = model.predict(X_test)\n",
        "y_pred = np.argmax(Y_pred, axis=1)\n",
        "\n",
        "print(y_true)\n",
        "print(y_pred)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[ 6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6\n",
            "  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6  6\n",
            "  6  6  6  6  6  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8\n",
            "  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8\n",
            "  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9 23 23 23 23 23\n",
            " 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23\n",
            " 23]\n",
            "[ 8  6  6  6  6  6  6  6  6  6  6  6  6  6  8  8  6  6  6  6  6  6  6  6\n",
            "  8  6  9  6  6  6  6  6  6  6  6  8  6  6  6  8  6  8  6  6  6  6  6  8\n",
            "  6  6  6  4  0  6  8  8  8  8  8  8  8  8 21  8  8  8  8  8  8  8  8  8\n",
            "  8  8  8  8  8  8  8  8  8  8  8  8  8  8  6  8  8  6  8  8  8  8  8  8\n",
            "  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  8  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  2  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  6  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  8  8  9  9  9  9  9  9  9  9  9  9  9  9\n",
            "  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9  9 23 21  3 23 23\n",
            " 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23\n",
            " 23]\n",
            "time: 1.22 s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s_9daykG0MAK",
        "colab_type": "text"
      },
      "source": [
        "## Membuat Confusion Matrix"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GhoUQvs1SYko",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 540
        },
        "outputId": "dce91d9e-87af-4600-b948-bc13bdeb1403"
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.utils.multiclass import unique_labels\n",
        "\n",
        "def plot_confusion_matrix(y_true, y_pred, classes,\n",
        "                          normalize=False,\n",
        "                          title=None,\n",
        "                          cmap=plt.cm.Blues):\n",
        "    \"\"\"\n",
        "    This function prints and plots the confusion matrix.\n",
        "    Normalization can be applied by setting `normalize=True`.\n",
        "    \"\"\"\n",
        "    if not title:\n",
        "        if normalize:\n",
        "            title = 'Normalized confusion matrix'\n",
        "        else:\n",
        "            title = 'Confusion matrix, without normalization'\n",
        "\n",
        "    # Compute confusion matrix\n",
        "    cm = confusion_matrix(y_true, y_pred)\n",
        "    # Only use the labels that appear in the data\n",
        "    #classes = classes[unique_labels(y_true, y_pred)]\n",
        "    if normalize:\n",
        "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
        "        print(\"Normalized confusion matrix\")\n",
        "    else:\n",
        "        print('Confusion matrix, without normalization')\n",
        "\n",
        "    print(cm)\n",
        "\n",
        "    fig, ax = plt.subplots(figsize=(5,5))\n",
        "    im = ax.imshow(cm, interpolation='nearest', cmap=cmap)\n",
        "    #ax.figure.colorbar(im, ax=ax)\n",
        "    # We want to show all ticks...\n",
        "    ax.set(xticks=np.arange(cm.shape[1]),\n",
        "           yticks=np.arange(cm.shape[0]),\n",
        "           # ... and label them with the respective list entries\n",
        "           xticklabels=classes, yticklabels=classes,\n",
        "           title=title,\n",
        "           ylabel='Label Benar',\n",
        "           xlabel='Label Prediksi')\n",
        "\n",
        "    # Rotate the tick labels and set their alignment.\n",
        "    plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\",\n",
        "             rotation_mode=\"anchor\")\n",
        "\n",
        "    # Loop over data dimensions and create text annotations.\n",
        "    fmt = '.2f' if normalize else 'd'\n",
        "    thresh = cm.max() / 2.\n",
        "    for i in range(cm.shape[0]):\n",
        "        for j in range(cm.shape[1]):\n",
        "            ax.text(j, i, format(cm[i, j], fmt),\n",
        "                    ha=\"center\", va=\"center\",\n",
        "                    color=\"white\" if cm[i, j] > thresh else \"black\")\n",
        "    fig.tight_layout()\n",
        "    return ax\n",
        "\n",
        "\n",
        "np.set_printoptions(precision=2)\n",
        "\n",
        "plot_confusion_matrix(y_true, y_pred, classes=labels, normalize=True,\n",
        "                      title='Normalized confusion matrix')"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Normalized confusion matrix\n",
            "[[ nan  nan  nan  nan  nan  nan  nan  nan  nan]\n",
            " [ nan  nan  nan  nan  nan  nan  nan  nan  nan]\n",
            " [ nan  nan  nan  nan  nan  nan  nan  nan  nan]\n",
            " [ nan  nan  nan  nan  nan  nan  nan  nan  nan]\n",
            " [0.02 0.   0.   0.02 0.79 0.15 0.02 0.   0.  ]\n",
            " [0.   0.   0.   0.   0.05 0.94 0.   0.02 0.  ]\n",
            " [0.   0.   0.   0.   0.   0.   0.99 0.   0.  ]\n",
            " [ nan  nan  nan  nan  nan  nan  nan  nan  nan]\n",
            " [0.   0.   0.03 0.   0.   0.   0.   0.03 0.93]]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:23: RuntimeWarning: invalid value encountered in true_divide\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fcdd732bbe0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "time: 558 ms\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "99nYLupr0eAs",
        "colab_type": "text"
      },
      "source": [
        "## Menyimpan dan Konversi Model ke \".tflite\""
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NxIkRiZJ0hVH",
        "colab_type": "text"
      },
      "source": [
        "Menyimpan model menggunakan tf.saved_model.save dan kemudian mengonversi model tersimpan ke format yang kompatibel tf lite."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DnpPEwsm_N10",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 173
        },
        "outputId": "f0c1f269-e63c-49b2-fc4a-429f9f95ae57"
      },
      "source": [
        "saved_model_dir = 'save/model'\n",
        "tf.saved_model.save(model, saved_model_dir)\n",
        "\n",
        "converter = tf.lite.TFLiteConverter.from_saved_model(saved_model_dir)\n",
        "tflite_model = converter.convert()\n",
        "\n",
        "with open('Plant_Leaf_Diseases_Classification_MobileNet.tflite', 'wb') as f:\n",
        "  f.write(tflite_model)"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Model.state_updates (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/tracking/tracking.py:111: Layer.updates (from tensorflow.python.keras.engine.base_layer) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "This property should not be used in TensorFlow 2.0, as updates are applied automatically.\n",
            "INFO:tensorflow:Assets written to: save/model/assets\n",
            "time: 14.9 s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LyEPqGw1ybSO",
        "colab_type": "text"
      },
      "source": [
        "Menyimpan model tflite versi kompatibel quantized"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S8A00bMzyHPF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "bd703fcb-7d08-4c8a-9f15-e009537da783"
      },
      "source": [
        "saved_model_dir = 'save/model'\n",
        "tf.saved_model.save(model, saved_model_dir)\n",
        "\n",
        "converter = tf.lite.TFLiteConverter.from_saved_model(saved_model_dir)\n",
        "converter.optimizations = [tf.lite.Optimize.DEFAULT]\n",
        "tflite_quant_model = converter.convert()\n",
        "\n",
        "with open('Plant_Leaf_Diseases_Classification_MobileNet_Quantized.tflite', 'wb') as f:\n",
        "  f.write(tflite_quant_model)"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Assets written to: save/model/assets\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Assets written to: save/model/assets\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "time: 15.1 s\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}