Copy_of_Copy_of_image_classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy_of_Copy_of_image_classification.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "pZoGDwnLbtKD"
      },
      "source": [
        "#import tensorflow\r\n",
        "#print(tensorflow.__version__)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NxlXB88PMXmI",
        "outputId": "6ea0aa27-8955-4af2-f1d8-1cf8042928f9"
      },
      "source": [
        "%cd /content/drive/MyDrive/ML Project\r\n"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/.shortcut-targets-by-id/1OGBxv0w8wrNxcfysVAcQP9R25afxB_kT/ML Project\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TkarOVV2dN1Y",
        "outputId": "4c63573e-b769-417d-b8e4-2cac3abbe9b8"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b-_DYSwKMxs7"
      },
      "source": [
        "!mkdir ~/.kaggle"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QHQ0nh84NpZH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6d3ab0a4-f8e3-4d16-e176-682402102680"
      },
      "source": [
        "!cp /content/kaggle.json ~/.kaggle"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "cp: cannot stat '/content/kaggle.json': No such file or directory\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "40BOal9XdM99"
      },
      "source": [
        ""
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HJ7CkSFxNtx3",
        "outputId": "f2cf6554-d237-41e2-cdde-7a361fd1fa48"
      },
      "source": [
        "!kaggle datasets download -d meowmeowmeowmeowmeow/gtsrb-german-traffic-sign"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Traceback (most recent call last):\n",
            "  File \"/usr/local/bin/kaggle\", line 5, in <module>\n",
            "    from kaggle.cli import main\n",
            "  File \"/usr/local/lib/python2.7/dist-packages/kaggle/__init__.py\", line 23, in <module>\n",
            "    api.authenticate()\n",
            "  File \"/usr/local/lib/python2.7/dist-packages/kaggle/api/kaggle_api_extended.py\", line 146, in authenticate\n",
            "    self.config_file, self.config_dir))\n",
            "IOError: Could not find kaggle.json. Make sure it's located in /root/.kaggle. Or use the environment method.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NH3-3bxTN4ng",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a29f1f57-8c6d-44b3-a073-c8b4d2e5f01a"
      },
      "source": [
        "!unzip gtsrb-german-traffic-sign.zip -d ."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "unzip:  cannot find or open gtsrb-german-traffic-sign.zip, gtsrb-german-traffic-sign.zip.zip or gtsrb-german-traffic-sign.zip.ZIP.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4LFjfIwpOCwX"
      },
      "source": [
        "# define some variables for later usage\r\n",
        "epochs = 10\r\n",
        "img_size = 150\r\n",
        "batch_size = 32"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ratw5bQrPhGX"
      },
      "source": [
        "data_dir = \"/content/drive/MyDrive/ML Project/Pneumonia\"\r\n",
        "TEST = '/content/drive/MyDrive/ML Project/Pneumonia/test/'\r\n",
        "TRAIN = '/content/drive/MyDrive/ML Project/Pneumonia/train/'\r\n",
        "VAL ='/content/drive/MyDrive/ML Project/Pneumonia/valid/'"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TNISOnxAPmyO",
        "outputId": "803dbe56-ab82-4321-b99c-8b561f495faa"
      },
      "source": [
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\r\n",
        "train_datagen = ImageDataGenerator(rescale = 1./255,\r\n",
        "                                   zoom_range = 0.2, \r\n",
        "                                   vertical_flip = True)\r\n",
        "\r\n",
        "# load a mini batch of images directly from the source folder then convert them into a vector of attributes\r\n",
        "training_set = train_datagen.flow_from_directory(directory = data_dir + '/train',\r\n",
        "                                                 batch_size = 8,\r\n",
        "                                                 target_size = (150, 150),\r\n",
        "                                                 shuffle = True,\r\n",
        "                                                 class_mode = 'binary')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 5216 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "smQb-0m4ibKy",
        "outputId": "2038ccc5-0e23-4bfe-aab8-397a8ec9c784"
      },
      "source": [
        "training_set.labels"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, ..., 1, 1, 1], dtype=int32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uwdLCz-jY-VZ",
        "outputId": "ab440955-dc03-4dcd-9843-0b04a9e93d84"
      },
      "source": [
        "test_datagen = ImageDataGenerator(rescale = 1./255)\r\n",
        "\r\n",
        "test_set = test_datagen.flow_from_directory(directory = data_dir + '/test',\r\n",
        "                                            batch_size = 8,\r\n",
        "                                            target_size = (150, 150),\r\n",
        "                                            shuffle = True,\r\n",
        "                                            class_mode = 'binary')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 624 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PGkmTNt3Pm-G"
      },
      "source": [
        "# import required libraries\r\n",
        "import os\r\n",
        "import cv2\r\n",
        "import numpy as np\r\n",
        "import matplotlib.pyplot as plt\r\n",
        "\r\n",
        "from keras.models import Sequential\r\n",
        "from keras.preprocessing import image\r\n",
        "from mlxtend.plotting import plot_confusion_matrix\r\n",
        "from keras.preprocessing.image import ImageDataGenerator\r\n",
        "from sklearn.metrics import accuracy_score, confusion_matrix\r\n",
        "\r\n",
        "from keras.layers import Dense, Conv2D, Dropout, Flatten, Activation\r\n",
        "from keras.layers import SeparableConv2D, MaxPool2D, BatchNormalization\r\n",
        "from keras.callbacks import ModelCheckpoint, ReduceLROnPlateau, EarlyStopping"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SlgvWxkGPp_-"
      },
      "source": [
        "model = Sequential()\r\n",
        "\r\n",
        "# 1st conv\r\n",
        "# 1st conv\r\n",
        "model.add(Conv2D(filters = 16, kernel_size = (3, 3), input_shape = (150, 150, 3), activation = 'relu', padding = 'same'))\r\n",
        "model.add(Conv2D(filters = 16, kernel_size = (3, 3), activation = 'relu', padding = 'same'))\r\n",
        "model.add(MaxPool2D(pool_size = (2, 2)))\r\n",
        "model.add(Flatten())\r\n",
        "model.add(Dense(32,  activation = 'relu'))\r\n",
        "\r\n",
        "model.add(Dense(units = 1, activation = 'sigmoid'))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iPpM4_ScVnQU"
      },
      "source": [
        "model.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "3tOE8nGbVqkE",
        "outputId": "a77ca472-34db-4919-ea69-e17264d2c66e"
      },
      "source": [
        "# fit the data to the cnn model\r\n",
        "cnn = model.fit_generator(training_set,\r\n",
        "                          epochs =5,\r\n",
        "                          steps_per_epoch = training_set.samples // 8,\r\n",
        "                          validation_data = test_set,\r\n",
        "                          validation_steps = test_set.samples // 8,\r\n",
        "                          )        "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:1844: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
            "  warnings.warn('`Model.fit_generator` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "652/652 [==============================] - ETA: 0s - loss: 0.4404 - accuracy: 0.8280"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "ERROR:root:Internal Python error in the inspect module.\n",
            "Below is the traceback from this internal error.\n",
            "\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Traceback (most recent call last):\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/IPython/core/interactiveshell.py\", line 2882, in run_code\n",
            "    exec(code_obj, self.user_global_ns, self.user_ns)\n",
            "  File \"<ipython-input-13-3fb31745922a>\", line 6, in <module>\n",
            "    validation_steps = test_set.samples // 8,\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py\", line 1861, in fit_generator\n",
            "    initial_epoch=initial_epoch)\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py\", line 1117, in fit\n",
            "    self._fit_frame = tf_inspect.currentframe()\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/utils/tf_inspect.py\", line 95, in currentframe\n",
            "    return _inspect.stack()[1][0]\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 1501, in stack\n",
            "    return getouterframes(sys._getframe(1), context)\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 1478, in getouterframes\n",
            "    frameinfo = (frame,) + getframeinfo(frame, context)\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 1448, in getframeinfo\n",
            "    filename = getsourcefile(frame) or getfile(frame)\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 696, in getsourcefile\n",
            "    if getattr(getmodule(object, filename), '__loader__', None) is not None:\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 725, in getmodule\n",
            "    file = getabsfile(object, _filename)\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 709, in getabsfile\n",
            "    return os.path.normcase(os.path.abspath(_filename))\n",
            "  File \"/usr/lib/python3.6/posixpath.py\", line 383, in abspath\n",
            "    cwd = os.getcwd()\n",
            "FileNotFoundError: [Errno 2] No such file or directory\n",
            "\n",
            "During handling of the above exception, another exception occurred:\n",
            "\n",
            "Traceback (most recent call last):\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/IPython/core/interactiveshell.py\", line 1823, in showtraceback\n",
            "    stb = value._render_traceback_()\n",
            "AttributeError: 'FileNotFoundError' object has no attribute '_render_traceback_'\n",
            "\n",
            "During handling of the above exception, another exception occurred:\n",
            "\n",
            "Traceback (most recent call last):\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/IPython/core/ultratb.py\", line 1132, in get_records\n",
            "    return _fixed_getinnerframes(etb, number_of_lines_of_context, tb_offset)\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/IPython/core/ultratb.py\", line 313, in wrapped\n",
            "    return f(*args, **kwargs)\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/IPython/core/ultratb.py\", line 358, in _fixed_getinnerframes\n",
            "    records = fix_frame_records_filenames(inspect.getinnerframes(etb, context))\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 1490, in getinnerframes\n",
            "    frameinfo = (tb.tb_frame,) + getframeinfo(tb, context)\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 1448, in getframeinfo\n",
            "    filename = getsourcefile(frame) or getfile(frame)\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 696, in getsourcefile\n",
            "    if getattr(getmodule(object, filename), '__loader__', None) is not None:\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 725, in getmodule\n",
            "    file = getabsfile(object, _filename)\n",
            "  File \"/usr/lib/python3.6/inspect.py\", line 709, in getabsfile\n",
            "    return os.path.normcase(os.path.abspath(_filename))\n",
            "  File \"/usr/lib/python3.6/posixpath.py\", line 383, in abspath\n",
            "    cwd = os.getcwd()\n",
            "FileNotFoundError: [Errno 2] No such file or directory\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "error",
          "ename": "FileNotFoundError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iDSF30qcV4iD",
        "outputId": "27a42a73-204d-430c-de7d-4fd13dda6934"
      },
      "source": [
        "model.predict_classes(test_set[30][0])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/sequential.py:450: UserWarning: `model.predict_classes()` is deprecated and will be removed after 2021-01-01. Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n",
            "  warnings.warn('`model.predict_classes()` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[1],\n",
              "       [1],\n",
              "       [0],\n",
              "       [1],\n",
              "       [0],\n",
              "       [1],\n",
              "       [1],\n",
              "       [1]], dtype=int32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LOSNkoktV9fU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f8fb5b1a-83e8-4059-f8bf-7e9275f2dbad"
      },
      "source": [
        "test_set.labels"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1], dtype=int32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 241
        },
        "id": "Ot_ZM37kczgy",
        "outputId": "2ff0ed55-48fd-439b-a12e-3e56299f4c1c"
      },
      "source": [
        "fig, ax = plt.subplots(1, 2, figsize = (10, 3))\r\n",
        "ax = ax.ravel()\r\n",
        "\r\n",
        "for i, met in enumerate(['accuracy', 'loss']):\r\n",
        "    ax[i].plot(cnn.history[met])\r\n",
        "    ax[i].plot(cnn.history['val_' + met])\r\n",
        "    ax[i].set_title('Model {}'.format(met))\r\n",
        "    ax[i].set_xlabel('epochs')\r\n",
        "    ax[i].set_ylabel(met)\r\n",
        "    ax[i].legend(['train', 'val'])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x216 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5TzScwenixt1"
      },
      "source": [
        "test_data = []\r\n",
        "test_labels = []\r\n",
        "\r\n",
        "for title in ['/NORMAL/', '/PNEUMONIA/']:\r\n",
        "        for img in (os.listdir(data_dir + '/test' + title)):\r\n",
        "            img = plt.imread(data_dir + '/test' + title + img)\r\n",
        "            img = cv2.resize(img, (img_size, img_size))\r\n",
        "            img = np.dstack([img, img, img])\r\n",
        "            img = img.astype('float32') / 255\r\n",
        "            if title == '/NORMAL/':\r\n",
        "                label = 0\r\n",
        "            elif title == '/PNEUMONIA/':\r\n",
        "                label = 1\r\n",
        "            test_data.append(img)\r\n",
        "            test_labels.append(label)\r\n",
        "        \r\n",
        "test_data = np.array(test_data)\r\n",
        "test_labels = np.array(test_labels)\r\n",
        "# get predictions\r\n",
        "preds = model.predict(test_data)\r\n",
        "\r\n",
        "acc = accuracy_score(test_labels, np.round(preds))*100\r\n",
        "# get the confusion matrix\r\n",
        "mat = confusion_matrix(test_labels, np.round(preds))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        },
        "id": "Ja3L9GPbqot6",
        "outputId": "c885d150-fa19-486b-c179-e81279f76897"
      },
      "source": [
        "from sklearn.metrics import ConfusionMatrixDisplay\r\n",
        "cmd = ConfusionMatrixDisplay(mat, display_labels=['Normal', 'Pneumonia'])\r\n",
        "cmd.plot(cmap = plt.cm.Blues)\r\n",
        "plt.title('Confusion Matrix')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Confusion Matrix')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWoAAAEWCAYAAABPON1ZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5wdVf3/8dd7NyG9B2KAxNCRmkAIBKQr0jSIiBRpooASQBEFxB8CflGqIEgRBUloCiK9S++QQEgIoUQJkAIphJhACCmf3x9zFm42W+5utszd+37mMY+dOXPmzLl3s5977pkzZxQRmJlZflW0dgXMzKxuDtRmZjnnQG1mlnMO1GZmOedAbWaWcw7UZmY550BtuSSpk6S7JM2TdMtKlHOwpAebsm6tQdJ9kg5r7XpY63CgtpUi6SBJYyQtkDQjBZSvNkHR+wH9gD4R8d3GFhIRN0TEbk1Qn+VI2klSSLqtWvrmKf2xIss5Q9L19eWLiD0iYlQjq2slzoHaGk3SicDFwO/IgupA4HJgRBMU/2XgzYhY0gRlNZdZwHBJfQrSDgPebKoTKOO/0zLn/wDWKJJ6AGcBx0bEvyLi44hYHBF3RcQvUp4Oki6WND0tF0vqkPbtJGmqpJ9Lmpla40ekfWcCpwPfSy31I6u3PCUNSi3Xdmn7cEn/lTRf0tuSDi5If6rguG0lvZi6VF6UtG3Bvsck/VbS06mcByX1reNt+Ay4HTggHV8JfA+4odp79UdJ70n6n6SxkrZP6bsDvyp4na8U1ONsSU8DnwBrp7Qfpv1XSLq1oPxzJT0sSUX/Aq2kOFBbYw0HOgK31ZHnNGAbYDCwOTAM+HXB/i8BPYA1gCOByyT1iojfkLXS/xERXSPi6roqIqkLcAmwR0R0A7YFxtWQrzdwT8rbB/gDcE+1FvFBwBHAasAqwEl1nRsYDRya1r8BvApMr5bnRbL3oDdwI3CLpI4RcX+117l5wTGHAEcB3YB3qpX3c2DT9CG0Pdl7d1h4Pog2y4HaGqsPMLueromDgbMiYmZEzALOJAtAVRan/Ysj4l5gAbBBI+uzDNhEUqeImBERE2vIsxfwVkRcFxFLIuIm4HXgmwV5/hYRb0bEQuBmsgBbq4h4BugtaQOygD26hjzXR8ScdM4LgQ7U/zqvjYiJ6ZjF1cr7hOx9/ANwPXBcREytpzwrYQ7U1lhzgL5VXQ+1WJ3lW4PvpLTPy6gW6D8Buja0IhHxMVmXwzHADEn3SNqwiPpU1WmNgu33G1Gf64CRwM7U8A1D0kmSJqXulo/IvkXU1aUC8F5dOyPieeC/gMg+UKwNc6C2xnoWWATsU0ee6WQXBasMZMVugWJ9DHQu2P5S4c6IeCAivg70J2sl/6WI+lTVaVoj61TlOuAnwL2ptfu51DXxS2B/oFdE9ATmkQVYgNq6K+rsxpB0LFnLfHoq39owB2prlIiYR3bB7zJJ+0jqLKm9pD0knZey3QT8WtKq6aLc6WRf1RtjHLCDpIHpQuapVTsk9ZM0IvVVLyLrQllWQxn3AuunIYXtJH0P2Ai4u5F1AiAi3gZ2JOuTr64bsIRshEg7SacD3Qv2fwAMasjIDknrA/8HfJ+sC+SXkursorHS5kBtjZb6W08ku0A4i+zr+kiykRCQBZMxwHhgAvBSSmvMuR4C/pHKGsvywbUi1WM68CFZ0PxxDWXMAfYmuxg3h6wlundEzG5MnaqV/VRE1PRt4QHgfrIhe+8An7J8t0bVzTxzJL1U33lSV9P1wLkR8UpEvEU2cuS6qhE11vbIF4rNzPLNLWozs5xzoDYzyzkHajOznHOgNjPLubpuVrBG6Nqzd/Tpv2ZrV8MaoGt7/xmUkulT32Xuh7NXal6Tyu5fjliysKi8sXDWAxGx+8qcb2X5f2gT69N/TU695s7WroY1wHYD6rtJ0PLkgD13WOkyYslCOmywf1F5Px13Wav/B3GgNrMyJCih2WMdqM2s/AioqGztWhTNgdrMylMJTd/tQG1mZchdH2Zm+ecWtZlZjgm3qM3M8k1uUZuZ5Z5HfZiZ5ZkvJpqZ5Ztw14eZWe65RW1mlmel1fVROjU1M2sqAiori1vqK0rqKOkFSa9ImijpzJR+raS3JY1Ly+CULkmXSJosabykLeo7h1vUZlaemq6PehGwS0QskNQeeErSfWnfLyLin9Xy7wGsl5atgSvSz1o5UJtZGWq6ro/InhC+IG22T0tdTw0fAYxOxz0nqaek/hExo7YD3PVhZuVJKm6BvpLGFCxHrViUKiWNA2YCD0XE82nX2al74yJJHVLaGsB7BYdPTWm1covazMpT8S3q2RExtK4MEbEUGCypJ3CbpE2AU4H3gVWAq4CTgbMaU1W3qM2s/BTbmm5gP3ZEfAQ8CuweETMiswj4GzAsZZsGDCg4bM2UVisHajMrTxWVxS31kLRqakkjqRPwdeB1Sf1TmoB9gFfTIXcCh6bRH9sA8+rqnwZ3fZhZWWrScdT9gVGSKskavzdHxN2SHpG0anYyxgHHpPz3AnsCk4FPgCPqO4EDtZmVpyYanhcR44EhNaTvUkv+AI5tyDkcqM2s/Hg+ajOzvCutW8gdqM2sPHk+ajOznPM0p2ZmOSZ3fZiZ5Z9b1GZm+SYHajOz/MqexOVAbWaWXxKqcKA2M8s1t6jNzHLOgdrMLOccqM3M8kxpKREO1GZWdoTcojYzy7uKCt+ZaGaWa25Rm5nlmfuozczyzy1qM7Mc88VEM7MS4FvIzczyTKXV9VE641PMzJqQpKKWIsrpKOkFSa9ImijpzJS+lqTnJU2W9A9Jq6T0Dml7cto/qL5zOFCbWVlqqkANLAJ2iYjNgcHA7pK2Ac4FLoqIdYG5wJEp/5HA3JR+UcpXJwdqMys7VRcTmyJQR2ZB2myflgB2Af6Z0kcB+6T1EWmbtH9X1XMiB2ozK08qcoG+ksYULEetUJRUKWkcMBN4CPgP8FFELElZpgJrpPU1gPcA0v55QJ+6quqLiWZWftSgW8hnR8TQujJExFJgsKSewG3AhitZw+W4RW1mZakJ+6g/FxEfAY8Cw4Gekqoaw2sC09L6NGBAqkM7oAcwp65yHajNrDwV3/VRdzHSqqkljaROwNeBSWQBe7+U7TDgjrR+Z9om7X8kIqKuc7jro4266br7eW3Cf+jarTMn/78jaswz+c13ue2fj7J06TK6dunEyBMPWKlzLlm8hBtG3cfU9z6gc5eOHHbkN+ndpwdvTJrC3bc/ydKlS6msrORb++7IehsMXKlz2RfemTqLX5134+fb09//kKMO/jpbbrY251x2Ows/XUT/1Xpx1kkH0LVzx1asab404Tjq/sAoSZVkjd+bI+JuSa8Bf5f0f8DLwNUp/9XAdZImAx8C9f7hlUyglhTAHyLi52n7JKBrRJzRgnV4DDgpIsa01Dkba9g2G/PVHYdw46h7a9y/8JNP+eff/83RI/ejV+/uzJ//cdFlfzhnHjeOvo+RP1v+/9dzz0ygU+eOnHbmD3lpzOvcddsTHPbDb9Klayd++ONv06NnV2ZMn8WfL72VM35/zEq9PvvCl9dclRsuOQGApUuXsdfhv2On4Rtzyu9v4IQf7MkWm67NnQ+9yPX/eoJjvr9bK9c2HxrTrVGbiBgPDKkh/b/AsBrSPwW+25BzlFLXxyJgX0l9G3NwQV9RWVhnvQF06VJ762nsi5PYbPD69OrdHYBu3bp8vm/M869x0bnXc/7vRnHzjQ+ybNmyos756vj/MGybjQHYfMj6vPXGu0QEaw7oR4+eXQH4Uv++LF68hCWLl9RVlDXSi69MZs3+fei/Wi/enT6LIZusBcDWg9fj0WdebeXa5Utz9FE3l1IK1EuAq4CfVd8haZCkRySNl/SwpIEp/VpJV0p6HjgvbV8h6TlJ/5W0k6RrJE2SdG1BeVekYTif32XU1syaOZdPPvmUP130dy78/XW8+NxEAD6YMYeXx77O8ScdyC9+dRgVEmNfmFRUmfM+mk/PXt0AqKysoGOnVfj444XL5Xnl5TdZY8BqtGtfVp+bLeahJ19htx02B2Dtgf14/LnXAPj30xP4YPZHrVm13FGFilryoNT+Wi4Dxks6r1r6pcCoiBgl6QfAJXwxuHxNYNuIWJqCcS+yK7LfIuvU3w74IfCipMERMQ44LSI+TH1OD0vaLH29qVEaV3kUQO9+qzfVa21Wy5YtY+q7H/DjE77L4sVL+OP5N/Lltfrz5hvvMPW9D/jDudcDsPizJXTt1hmAa/58O3PmzGPpkqXMnTuf83+XjdnfYect2Hr4pvWec8b02dx9+xMcc1yDvvVZkRYvXsITz0/iJ4fuDsD/O34/LrzqLq75xyNsv/VXaNeu1P7cm1deWsvFKKnfXET8T9Jo4HigsKk2HNg3rV8HFAbyW9IYxyp3RURImgB8EBETACRNBAYB44D9U/BtR3ahYCOg1kAdEVeRtfb58lc2q/PqbV706NmNzl060aHDKnTosArrrLsm06fNIgK22npj9t5nhxWO+cHR2WdfbX3UPXp246O5Wat66dJlfLrwM7p06QTAR3Pn87er7uCgw/ak76o9m/8FlqFnxr7BhuusQZ/0rWbQgNW49LfZXcvvTJvF0y++3prVyxdPytTsLia7V75LfRmT6lfJFqWfywrWq7bbSVoLOAnYNSI2A+4B2tyl8k03W5e3/zONpUuX8dlni3lnygz6fak36284kFdefvPzi4sff7yQD+fMK6rMTTZbhxdSF8orL7/JuhsMQBILP/mUv1z+L/YesT1rr7NGPaVYYz34xCvstuPmn29/+FF2V/OyZcu45h+PsO8eW7dW1XJHgFTckgcl1aIGSF0SN5MF62tS8jNkQ1yuAw4GnlyJU3QnC+7zJPUD9gAeW4nyWsXoa+5m8pvv8fGChZzxqyvZfa/tWLo0+2Kx3Q6D6de/DxtuNIjzz74WSWyz3Wb0X31VAPb85le58tJ/EsuCyspKvnPArvTu06Pec2697abccO29nP2bv9K5c0cOOXJvAJ58/GVmz5rLA/c9ywP3PQvAMcftt9wFTFs5Cz/9jOfHTebUY/f9PO3BJ8Zxyz3PAbDz8I355tfqvLmuzOTnQmExSi5QJxcCIwu2jwP+JukXwCyg5oHDRYiIVyS9DLxOdj/+0ytT0dZy6A/2rjfPLl8fxi5fX2H0EEOGbsiQobXfAdu7T48Vuj0A2rdvx+E/+tYK6bvtMZzd9hheb32s8Tp1XIV/33j6cmkHfOurHPCtr7ZSjfKvIicXCotRMoE6IroWrH8AdC7Yfodspqrqxxxe23ZETAE2qWXfcscVpO/U4IqbWf7kqFujGCUTqM3Mmopwi9rMLPfcojYzyzlfTDQzyzP3UZuZ5ZtQQx4c0OocqM2sLLlFbWaWc+6jNjPLM/dRm5nlWzbXR+lEagdqMytLJRSnHajNrDz5zkQzszzzfNRmZvnWlPNRSxog6VFJr6XH952Q0s+QNE3SuLTsWXDMqZImS3pD0jfqO4db1GZWhpp0PuolwM8j4iVJ3YCxkh5K+y6KiAuWO7O0Edn8+RsDqwP/lrR+tSdRLcctajMrS03Voo6IGRHxUlqfD0wC6nqU0Qjg7xGxKCLeBiYDK04MX8CB2szKj7KLicUsDSpWGgQMAZ5PSSMljZd0jaReKW0NsoeSVJlK3YHdgdrMyk/VOOpiFqCvpDEFy1E1lil1BW4FfhoR/wOuANYBBgMzyJ5M1SjuozazstSAPurZEVHnAycltScL0jdExL/g8ydRVe3/C3B32pwGDCg4fM2UViu3qM2sLDXhqA8BVwOTIuIPBen9C7J9G3g1rd8JHCCpg6S1gPWAF+o6h1vUZlaWmnDUx3bAIcAESeNS2q+AAyUNBgKYAhwNEBETJd0MvEY2YuTYukZ8gAO1mZWjJpyUKSKeykpcwb11HHM2cHax53CgNrOykz04oHTuTHSgNrOyVFFCt5A7UJtZWSqhOO1AbWblRyU2KZMDtZmVpRLqoq49UEu6lGxYSY0i4vhmqZGZWQtoKxcTx7RYLczMWpDIRn6UiloDdUSMKtyW1DkiPmn+KpmZNb8SalDXfwu5pOGSXgNeT9ubS7q82WtmZtZcipyQKS8XHIuZ6+Ni4BvAHICIeAXYoTkrZWbW3Jpqro+WUNSoj4h4r9onS533pZuZ5Zloeze8vCdpWyDSVH4nkD3BwMysZJXSqI9iuj6OAY4lewLBdLJJsI9tzkqZmTWnYrs98tLorrdFHRGzgYNboC5mZi2mlLo+ihn1sbakuyTNkjRT0h2S1m6JypmZNRcVueRBMV0fNwI3A/3JHm1+C3BTc1bKzKy5tbXheZ0j4rqIWJKW64GOzV0xM7Pmko36KG7Jg7rm+uidVu+TdArwd7K5P75HHU8uMDPLPbWdBweMJQvMVa/m6IJ9AZzaXJUyM2tueenWKEZdc32s1ZIVMTNrKVVdH6WiqDsTJW0CbERB33REjG6uSpmZNbdSalEXMzzvN8CladkZOA/4VjPXy8ysWTXV8DxJAyQ9Kuk1SRMlnZDSe0t6SNJb6WevlC5Jl0iaLGm8pC3qO0cxoz72A3YF3o+II4DNgR5FHGdmlksSVFaoqKUIS4CfR8RGwDbAsZI2Ak4BHo6I9YCH0zbAHsB6aTkKuKK+ExQTqBdGxDJgiaTuwExgQDG1NzPLq6YaRx0RMyLipbQ+n2wupDWAEUDVvP6jgH3S+ghgdGSeA3pK6l/XOYrpox4jqSfwF7KRIAuAZ4s4zswstxrQRd1XUuETr66KiKtqLlODgCHA80C/iJiRdr0P9EvrawDvFRw2NaXNoBbFzPXxk7R6paT7ge4RMb6+48zM8kqoIXN9zI6IofWWKXUFbgV+GhH/K2yNR0RIqvUZtPWp64aXWju4JW1R1dQ3Mys5TTwzXpoC+lbghoj4V0r+QFL/iJiRujZmpvRpLN99vGZKq1VdLeoL69gXwC511rxM9e28CodvNai1q2EN0Gurka1dBWuARZOnNkk5TTU8T1lBVwOTIuIPBbvuBA4Dzkk/7yhIHynp78DWwLyCLpIa1XXDy84rUXczs9wSUNl0TertgEOACZLGpbRfkQXomyUdCbwD7J/23QvsCUwGPgGOqO8ERd3wYmbW1jTVnYkR8RS1D7netYb8QQMfvuJAbWZlqc3dQm5m1pZkj9kqnUhdzC3kkvR9Saen7YGShjV/1czMmk8pzUddzJ2JlwPDgQPT9nzgsmarkZlZC2hTD7cFto6ILSS9DBARcyWt0sz1MjNrNgLa5SUKF6GYQL1YUiXZ2GkkrQosa9ZamZk1sxKK00UF6kuA24DVJJ1NNpver5u1VmZmzUhq0C3kra6YuT5ukDSWbDyggH0iYlKz18zMrBmVUJyuP1BLGkh298xdhWkR8W5zVszMrDnlZURHMYrp+riHLx5y2xFYC3gD2LgZ62Vm1mwExT4UIBeK6frYtHA7zar3k1qym5nlX47GSBejwXcmRsRLkrZujsqYmbUUFfVExHwopo/6xILNCmALYHqz1cjMrJmJttei7lawvoSsz/rW5qmOmVnLaDOBOt3o0i0iTmqh+piZtYhSmpSprkdxtYuIJZK2a8kKmZk1Nwkqi5npKCfqalG/QNYfPU7SncAtwMdVOwueC2ZmVnLa1J2JZGOn55A9I7FqPHUADtRmVpLa0sXE1dKIj1f5IkBXafRjz83M8qCEGtR1BupKoCs1PwvMgdrMSpioaCPjqGdExFktVhMzsxYiSqtFXdd1zxJ6GWZmDSBoV6GilnqLkq6RNFPSqwVpZ0iaJmlcWvYs2HeqpMmS3pD0jWKqW1egXuEx52ZmbUFVi7qJHsV1LbB7DekXRcTgtNwLIGkj4ACySe12By5P96vUqdZAHREfFlVFM7MSVJEeHlDfUp+IeAIoNl6OAP4eEYsi4m1gMlDvw8JLaMi3mVnTaUCLuq+kMQXLUUWeYqSk8alrpFdKWwN4ryDP1JRWJwdqMys7Igt+xSzA7IgYWrBcVcQprgDWAQYDM4ALV6a+DZ7m1Mys5Kl570yMiA8+P5X0F+DutDkNGFCQdc2UVie3qM2s7GR3JjZNH3WN5Uv9Cza/TXbjIMCdwAGSOkhaC1iPbLqOOrlFbWZlqana05JuAnYi68ueCvwG2EnSYLKbA6cARwNExERJNwOvkU0bfWxELK3vHA7UZlaWmqrnIyIOrCH56jrynw2c3ZBzOFCbWRlS25iP2sysraoa9VEqHKjNrCy1tfmozczaFrWRR3GZmbVV7vowMysBblGbmeVc6YRpB2ozK0MCKt2iNjPLtxKK0w7UZlaOhEqo88OB2szKklvUZmY5lg3PK51I7UBtZuWn+Och5oIDtZmVJd9CbmaWY9mDA1q7FsVzoDazsuRRH2ZmOVdCPR8O1LaikWddzwNPvUrfXt149h+nAXD7v1/i3Kvu5Y0pH/DwtScxZKMvt3It25YOq7Tjnqt+Sof27ahsV8mdD7/MOVfdu1yeNfv14vIzDqFHt05UVlRw5p/u4KFnXlup8w5cvQ9Xn30EvXt0Ydzr73LM6aNZvGQpPzloFw4ZMZylS5cx+6MFHHfW9bz3/tyVOlfelFKLutkmkJK0VNI4Sa9KukVS5+Y6V1OSNFTSJa1dj9Z04N7b8M9Ljl0u7SvrrM7o837EtkPWaaVatW2LPlvCiB9fwvYHn8MOB/2eXYdvxNBNBi2X5+dH7s7t/36JHb9/Lkee9jcuOPl7RZd/4N5bc/KP9lwh/YyRI7jixkfZct8zmfe/hRwyYjgA4994j10OPY+vHvR77nz4Zc44fp+Ven15U9VHXcySB80509/CiBgcEZsAnwHHNOO5mkxEjImI41u7Hq1puy3WpVf35T9XN1jrS6w3qF8r1ag8fLzwMwDat6ukfbtKImL5DBF069IRgO5dO/H+7HkAVFSIs47fh4dH/YKnbjyVw7+9XdHn3GGr9bnjkZcBuOme59lzx80BeGrsWyxctBiAFydMYY3Veq7Ua8udIp9AnpeRIS01JeuTwLqSdpL0mKR/Snpd0g1Kcw1K2lLS45LGSnqg6nHrKf/QtN5X0pS0frik2yU9JGmKpJGSTpT0sqTnJPVO+Qan7fGSbpPUq6DccyW9IOlNSdun9J0k3Z3Wh0l6NpX5jKQNWuj9sjJUUSGeuOEU3nzwHB57/nXGTnxnuf3nXHUv++8xjFfv/i03X/xjfnn+LQAcMmJb5i1YyK6Hnc8uh53Poftsx8DV+9R7vt49ujBv/kKWLl0GwPSZc1l9tR4r5DtkxPCV7mLJIxW51FuOdI2kmZJeLUjrnWLTW+lnVdyRpEskTU4xaYti6trsfdSS2gF7APenpCHAxsB04GlgO0nPA5cCIyJilqTvkT2l9wf1FL9JKq8jMBk4OSKGSLoIOBS4GBgNHBcRj0s6i+xR7j9Nx7eLiGGS9kzpX6tW/uvA9hGxRNLXgN8B36nhNR4FHAUwYODAYt4WsxUsWxbscPA5dO/aievP/xFfWac/k/4z4/P93/nGUG68+zkuu+ERttp0La4881C2PeB37Lz1hmy87hqM2HUIAN27dGSdAasy/+NPuePy4wDo1b0z7du3Y6+dNgPgmNNHf94ir8v+e2zF4K8MZO+j/9gMr7j1ZF0fTdZavhb4E1msqXIK8HBEnCPplLR9MlksXC8tWwNXpJ91as5A3UnSuLT+JNnj07cFXoiIqQBp/yDgI7Kg+1BqYFcCM6oXWINHI2I+MF/SPOCulD4B2ExSD6BnRDye0kcBtxQc/6/0c2yqR3U9gFGS1gMCaF9TJSLiKuAqgC23HBo15TEr1v8WLOTJsW+y6/CNlgvU3x8xnO8efxkAL054m44d2tOnZxckcfIFt/DIc5NWKGuHg88Bsj7qgf37cO5flr9A2aNbJyorK1i6dBmrr9aL6TO/CN47DtuAE4/4BnsffTGfLV7SHC+1VTVVmI6IJyQNqpY8AtgprY8CHiML1COA0ZH1az0nqaek/hFRZ7xriT7qwRFxXER8ltIXFeRZSvZhIWBiQf5NI2K3lGdJQT07VjtHYVnLCraXUdyHUFX+qnpU91uyD4NNgG/WcH6zJtGnZ1e6d+0EQMcO7dl52Ia8NeWD5fJMe/9Ddtgq631bf1A/OqzSntlzF/DIc5P4wXe+SrvK7M9knYGr0bnjKkWd98kxbzJil6wlfuBeW3PfE+MB2HT9Nbno1AM46Od/ZvbcBU3yGnOn+L6PvpLGFCxHFVF6v4Lg+z5QdYFnDeC9gnxTU1qd8jI87w1gVUnDI+JZSe2B9SNiIjAF2BJ4AdivIYVGxDxJcyVtHxFPAocAj9d3XIEewLS0fnhDzl3Kjjztbzw99i3mfLSAjff6NacctSe9unfh5AtuYfbcBXzvZ1ey6fprcOulI1u7qm3Gl/p25/IzDqGyooKKCnHbv1/igade5dSj92LcpHe574kJ/Pri2/jjaQfykwN3JoBjz7wOgNG3P8PA/r15/PpTkGD23AV8/6SrijrvGX+6g6vPPoLTfrw34994j+vueBaAs07Yhy6dOnDtOUcCMPX9uRz08z83y2tvLQ3o+pgdEUMbe56ICEkr9U07F4E6Ij6TtB9wSequaEfWvzwRuAC4OX2K3dOI4g8DrkzDA/8LHNGAY88j6/r4dSPPXZKuPrvmt2jvnTdv4ZqUj4mTp7Pj989dIf33f/7iv90bb7/P7j+8aIU8EcFvL7+L315+1wr7qtx09/M1pr8zbQ5fO/yCFdK/feyfiql2SWvm8RwfVHVppIERM1P6NGBAQb41+aIxWCutMATIVsqWWw6Np58f09rVsAbotZW/GZSSRW/czLJPZq5UnP3KpkNi9J2PFZV32No9x9bXok591HenblIknQ/MKbiY2DsifilpL2AksCfZRcRLImJYfXXIRYvazKwlZd3PTdOmlnQT2YXDvpKmko0gO4esJ+BI4B1g/5T9XrIgPRn4hCK/4TtQm1n5acL5qCPiwFp27VpD3gCOrSFvnRyozaws5eOew+I4UJtZGRLKye3hxXCgNrOyVEJx2oHazMpPsfN45IUDtZmVpxKK1A7UZlaWSunBAQ7UZlaW3EdtZpZnTTiOuszj4YcAAAifSURBVCU4UJtZWXLXh5lZjgm3qM3Mcq+E4rQDtZmVqRKK1A7UZlaW8vKE8WI4UJtZWSqdMO1AbWblqoQitQO1mZWdpnxwQEtwoDaz8uMbXszM8q+E4rQDtZmVIz84wMws90ooTjtQm1n58YMDzMxKQRNGaklTgPnAUmBJRAyV1Bv4BzAImALsHxFzG1N+RdNU08ystKjIfw2wc0QMjoihafsU4OGIWA94OG03igO1mZUlqbhlJYwARqX1UcA+jS3IgdrMyo+gosgF6CtpTMFyVA0lBvCgpLEF+/tFxIy0/j7Qr7HVdR+1mZWpopvLswu6M2rz1YiYJmk14CFJrxfujIiQFI2pJbhFbWZlqOrBAU3V9RER09LPmcBtwDDgA0n9AdLPmY2trwO1mZUlFbnUW47URVK3qnVgN+BV4E7gsJTtMOCOxtbVXR9mVpaa8IaXfsBt6U7HdsCNEXG/pBeBmyUdCbwD7N/YEzhQm1lZaqpbyCPiv8DmNaTPAXZtinM4UJtZWfKdiWZmOdYEY6RblAO1mZUlPzjAzCzvSidOO1CbWXkqoTjtQG1m5UhUlFAntQO1mZWdqjsTS4XvTDQzyzm3qM2sLJVSi9qB2szKkofnmZnlmW94MTPLt1K7mOhAbWZlyV0fZmY55xa1mVnOlVCcdqA2szJVQpHagdrMyo6gpG4hV0SjH4xrNZA0i+yxO21NX2B2a1fCGqSt/s6+HBGrrkwBku4ne3+KMTsidl+Z860sB2oriqQxETG0tethxfPvrO3wXB9mZjnnQG1mlnMO1Fasq1q7AtZg/p21Ee6jNjPLObeozcxyzoHazCznHKjLgKSQdGHB9kmSzmjhOjwmqeyHiklaKmmcpFcl3SKpc2vXqRiShkq6pLXrUa4cqMvDImBfScUO8F+OJN/B2nQWRsTgiNgE+Aw4prUrVIyIGBMRx7d2PcqVA3V5WEI2AuBn1XdIGiTpEUnjJT0saWBKv1bSlZKeB85L21dIek7SfyXtJOkaSZMkXVtQ3hWSxkiaKOnMlnqBJepJYN30Xj4m6Z+SXpd0g5Td3yxpS0mPSxor6QFJ/VP6599QJPWVNCWtHy7pdkkPSZoiaaSkEyW9nH53vVO+wWl7vKTbJPUqKPdcSS9IelPS9il9J0l3p/Vhkp5NZT4jaYOWfuPKjQN1+bgMOFhSj2rplwKjImIz4Aag8OvtmsC2EXFi2u4FDCcL+HcCFwEbA5tKGpzynJbuhtsM2FHSZs3yakpc+payBzAhJQ0BfgpsBKwNbCepPdnvZ7+I2BK4Bji7iOI3AfYFtkr5P4mIIcCzwKEpz2jg5PR7nwD8puD4dhExLNWnML3K68D2qczTgd8V9aKt0fyVtkxExP8kjQaOBxYW7BpO9kcNcB1wXsG+WyJiacH2XRERkiYAH0TEBABJE4FBwDhgf0lHkf3f6k8WeMY3w0sqVZ0kjUvrTwJXA9sCL0TEVIC0fxDwEVnQfSg1sCuBGUWc49GImA/MlzQPuCulTwA2Sx/WPSPi8ZQ+Cril4Ph/pZ9jUz2q6wGMkrQeEED7IupkK8GBurxcDLwE/K3I/B9X216Ufi4rWK/abidpLeAkYKuImJu6RDo2vrpt0sKIGFyYkIJw4fu5lOxvU8DEiBheQzlL+OIbcfX3uPrvpvD3VszffFX+qnpU91uyD4NvSxoEPFZEmbYS3PVRRiLiQ+Bm4MiC5GeAA9L6wWStvMbqThbc50nqR/bV3hrvDWBVScMBJLWXtHHaNwXYMq3v15BCI2IeMLeq/xk4BHi8jkOq6wFMS+uHN+Tc1jgO1OXnQpaf3vE44AhJ48n+YE9obMER8QrwMlkf5o3A0ytRz7IXEZ+RBeFzJb1C1rW0bdp9AfBjSS9T/HSdhQ4Dzk+/98HAWQ049jzg9+nc/lbeAnwLuZlZzrlFbWaWcw7UZmY550BtZpZzDtRmZjnnQG1mlnMO1NaimnL2uDT/yH5p/a+SNqoj706Stq1tfx3HTalpMqva0qvlWdDAc50h6aSG1tHaPgdqa2l1zh7X2Jn6IuKHEfFaHVl24osxyGYlxYHaWlPh7HFPSroTeE1SpaTzJb2YZnc7GkCZP0l6Q9K/gdWqCqo2m9zukl6S9IqyGQEHkX0g/Cy15reXtKqkW9M5XpS0XTq2j6QH0+x/fyW7jbtOaba6semYo6rtuyilPyxp1ZS2jqT70zFPStqwKd5Ma7t8V5G1ioLZ4+5PSVsAm0TE2ynYzYuIrSR1AJ6W9CDZDHMbkE301A94jWxGucJyVwX+AuyQyuodER9KuhJYEBEXpHw3AhdFxFPKpnZ9APgK2WxxT0XEWZL2Yvnb7Wvzg3SOTsCLkm6NiDlAF2BMRPxM0ump7JFkU84eExFvSdoauBzYpRFvo5UJB2praXXNHvd2St+NbJa3qjksegDrATsAN6UZ/aZLeqSG8rcBnqgqK81vUpOvARulCZEAukvqms6xbzr2Hklzi3hNx0v6dlofkOo6h2wSpH+k9OuBf6VzbAvcUnDuDkWcw8qYA7W1tNpmjyucqU/AcRHxQLV8ezZhPSqAbSLi0xrqUjRJO5EF/eER8Ymkx6h9xsBI5/2o+ntgVhf3UVsePUA24VB7AEnrS+oCPAF8L/Vh9wd2ruHY54Ad0pSrKD3RBJgPdCvI9yDZhFSkfFWB8wngoJS2B9nDEurSA5ibgvSGZC36KhV8MbPdQWRdKv8D3pb03XQOSdq8nnNYmXOgtjz6K1n/80uSXgX+TPbt7zbgrbRvNNkTS5YTEbOAo8i6GV7hi66Hu4BvV11MJHuAwtB0sfI1vhh9ciZZoJ9I1gXybj11vZ9sLu5JwDlkHxRVPgaGpdewC1/MUHcwcGSq30RgRBHviZUxz55nZpZzblGbmeWcA7WZWc45UJuZ5ZwDtZlZzjlQm5nlnAO1mVnOOVCbmeXc/wfRoXZ0OPc+vQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cFK20JT2qyZK",
        "outputId": "8ab28f1d-486d-4a21-da54-2554e4f5e4b0"
      },
      "source": [
        "# calculate precision and recall\r\n",
        "tn, fp, fn, tp = mat.ravel()\r\n",
        "print('Test Metrics:')\r\n",
        "precision = tp / (tp + fp) * 100\r\n",
        "recall = tp / (tp + fn) * 100\r\n",
        "print('Accuracy: {}%'.format(acc))\r\n",
        "print('Precision: {}%'.format(precision))\r\n",
        "print('Recall: {}%'.format(recall))\r\n",
        "print('F1-score: {}'.format(2*precision*recall/(precision+recall)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test Metrics:\n",
            "Accuracy: 85.57692307692307%\n",
            "Precision: 82.75109170305677%\n",
            "Recall: 97.17948717948718%\n",
            "F1-score: 89.38679245283019\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zZeDrne8q2NM",
        "outputId": "4dcd9ac7-b8b2-46b3-a9f4-35581f3a7a53"
      },
      "source": [
        "print('Train Metric:')\r\n",
        "print('Train accuracy: {}%'.format(np.round(cnn.history['accuracy'][-1]*100, 2)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train Metric:\n",
            "Train accuracy: 92.06%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M2_5VLktq7Zq"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}