mumpy_CNN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Yuasset/Assets/blob/master/mumpy_CNN.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "KfbJ5p3G1H1h",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "965419fb-c4b6-4414-ddb8-54f7c5ee81c4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cloning into 'urban-sounds-files'...\n",
            "remote: Enumerating objects: 12, done.\u001b[K\n",
            "remote: Counting objects: 100% (3/3), done.\u001b[K\n",
            "remote: Total 12 (delta 0), reused 3 (delta 0), pack-reused 9\u001b[K\n",
            "Unpacking objects: 100% (12/12), done.\n",
            "Checking out files: 100% (8/8), done.\n"
          ]
        }
      ],
      "source": [
        "#Model hazırlanması ve eğitimi\n",
        "! git clone https://github.com/Karamubarek/urban-sounds-files.git"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "feffO8z6CVga"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "a = np.load('urban-sounds-files/X_train.npz', allow_pickle=True)\n",
        "X_train = a['data']\n",
        "a = np.load('urban-sounds-files/X_test.npz', allow_pickle=True)\n",
        "X_test = a['data'] \n",
        "a = np.load('urban-sounds-files/X_val.npz', allow_pickle=True)\n",
        "X_val = a['data']\n",
        "a = np.load('urban-sounds-files/y_train.npz', allow_pickle=True)\n",
        "y_train = a['data']\n",
        "a = np.load('urban-sounds-files/y_test.npz', allow_pickle=True)\n",
        "y_test = a['data']\n",
        "a = np.load('urban-sounds-files/y_val.npz', allow_pickle=True)\n",
        "y_val = a['data']"
      ],
      "metadata": {
        "id": "gWbtehWFYs6A"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#CNN modeli hazırlayın\n",
        "#labelencoder = LabelEncoder()\n",
        "#dataset['diagnosis]']= labelencoder.fit_transform(dataset['diagnosis'].values)\n",
        "\n",
        "model = tf.keras.Sequential()\n",
        "\n",
        "model.add(tf.keras.layers.Conv2D(64,\n",
        "                                 kernel_size = (3,3),\n",
        "                                 strides = (1,1),\n",
        "                                 padding = 'same',\n",
        "                                 activation = 'relu',\n",
        "                                 input_shape = (128,128,1)))\n",
        "                                  \n",
        "\n",
        "model.add(tf.keras.layers.MaxPool2D(2,2))\n",
        "\n",
        "model.add(tf.keras.layers.Conv2D(64,\n",
        "                                 kernel_size = (3,3),\n",
        "                                 strides = (1,1),\n",
        "                                 padding = 'same',\n",
        "                                 activation = 'relu',\n",
        "                                 ))\n",
        "model.add(tf.keras.layers.MaxPool2D(2,2))\n",
        "model.add(tf.keras.layers.Conv2D(64,\n",
        "                                 kernel_size = (3,3),\n",
        "                                 strides = (1,1),\n",
        "                                 padding = 'same',\n",
        "                                 activation = 'relu',\n",
        "                                 ))\n",
        "\n",
        "model.add(tf.keras.layers.Flatten())\n",
        "\n",
        "model.add(tf.keras.layers.Dense(64, activation= 'relu'))\n",
        "model.add(tf.keras.layers.Dropout(0.5))\n",
        "\n",
        "\n",
        "model.add(tf.keras.layers.Dense(64, activation= 'relu'))\n",
        "model.add(tf.keras.layers.Dropout(0.5))\n",
        "\n",
        "model.add(tf.keras.layers.Dense(10,activation='softmax'))\n",
        "\n",
        "model.compile(optimizer='adam',\n",
        "              loss='sparse_categorical_crossentropy',\n",
        "              metrics=['accuracy']\n",
        "              )"
      ],
      "metadata": {
        "id": "zqZEQyfzBzdP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "results= model.fit(X_train, y_train,\n",
        "                   batch_size=128,\n",
        "                   epochs=50,\n",
        "                   validation_data= (X_val, y_val)\n",
        "                   )"
      ],
      "metadata": {
        "id": "VgzMv-HrBdsn",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "58066d5b-5077-49b4-d7dd-38fba34c8ab0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "55/55 [==============================] - 16s 120ms/step - loss: 2.1024 - accuracy: 0.2215 - val_loss: 1.6960 - val_accuracy: 0.3975\n",
            "Epoch 2/50\n",
            "55/55 [==============================] - 5s 89ms/step - loss: 1.7464 - accuracy: 0.3658 - val_loss: 1.2963 - val_accuracy: 0.5590\n",
            "Epoch 3/50\n",
            "55/55 [==============================] - 5s 89ms/step - loss: 1.4436 - accuracy: 0.4918 - val_loss: 0.9229 - val_accuracy: 0.7400\n",
            "Epoch 4/50\n",
            "55/55 [==============================] - 5s 89ms/step - loss: 1.2162 - accuracy: 0.5847 - val_loss: 0.7619 - val_accuracy: 0.7858\n",
            "Epoch 5/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 1.0477 - accuracy: 0.6394 - val_loss: 0.6712 - val_accuracy: 0.8121\n",
            "Epoch 6/50\n",
            "55/55 [==============================] - 5s 90ms/step - loss: 0.9114 - accuracy: 0.6898 - val_loss: 0.6088 - val_accuracy: 0.8225\n",
            "Epoch 7/50\n",
            "55/55 [==============================] - 5s 90ms/step - loss: 0.8344 - accuracy: 0.7181 - val_loss: 0.5913 - val_accuracy: 0.8373\n",
            "Epoch 8/50\n",
            "55/55 [==============================] - 5s 90ms/step - loss: 0.7633 - accuracy: 0.7343 - val_loss: 0.5422 - val_accuracy: 0.8465\n",
            "Epoch 9/50\n",
            "55/55 [==============================] - 5s 91ms/step - loss: 0.6939 - accuracy: 0.7639 - val_loss: 0.4975 - val_accuracy: 0.8557\n",
            "Epoch 10/50\n",
            "55/55 [==============================] - 5s 91ms/step - loss: 0.6115 - accuracy: 0.7881 - val_loss: 0.4870 - val_accuracy: 0.8660\n",
            "Epoch 11/50\n",
            "55/55 [==============================] - 5s 91ms/step - loss: 0.5647 - accuracy: 0.8020 - val_loss: 0.4392 - val_accuracy: 0.8774\n",
            "Epoch 12/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.5129 - accuracy: 0.8208 - val_loss: 0.4530 - val_accuracy: 0.8786\n",
            "Epoch 13/50\n",
            "55/55 [==============================] - 5s 91ms/step - loss: 0.4966 - accuracy: 0.8318 - val_loss: 0.4695 - val_accuracy: 0.8740\n",
            "Epoch 14/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.4530 - accuracy: 0.8401 - val_loss: 0.4591 - val_accuracy: 0.8545\n",
            "Epoch 15/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.4515 - accuracy: 0.8392 - val_loss: 0.4308 - val_accuracy: 0.8877\n",
            "Epoch 16/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.4150 - accuracy: 0.8504 - val_loss: 0.4141 - val_accuracy: 0.8877\n",
            "Epoch 17/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.3774 - accuracy: 0.8628 - val_loss: 0.4479 - val_accuracy: 0.8877\n",
            "Epoch 18/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.3861 - accuracy: 0.8664 - val_loss: 0.4643 - val_accuracy: 0.8855\n",
            "Epoch 19/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.3802 - accuracy: 0.8601 - val_loss: 0.4317 - val_accuracy: 0.8946\n",
            "Epoch 20/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.3475 - accuracy: 0.8752 - val_loss: 0.4665 - val_accuracy: 0.8751\n",
            "Epoch 21/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.3393 - accuracy: 0.8807 - val_loss: 0.4365 - val_accuracy: 0.8923\n",
            "Epoch 22/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.3144 - accuracy: 0.8873 - val_loss: 0.4812 - val_accuracy: 0.8889\n",
            "Epoch 23/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.3115 - accuracy: 0.8890 - val_loss: 0.4333 - val_accuracy: 0.8969\n",
            "Epoch 24/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2984 - accuracy: 0.8956 - val_loss: 0.4684 - val_accuracy: 0.8958\n",
            "Epoch 25/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2981 - accuracy: 0.8909 - val_loss: 0.4688 - val_accuracy: 0.8809\n",
            "Epoch 26/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2901 - accuracy: 0.8928 - val_loss: 0.4747 - val_accuracy: 0.8981\n",
            "Epoch 27/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2703 - accuracy: 0.9031 - val_loss: 0.5173 - val_accuracy: 0.8900\n",
            "Epoch 28/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2904 - accuracy: 0.8953 - val_loss: 0.4897 - val_accuracy: 0.8900\n",
            "Epoch 29/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2794 - accuracy: 0.8991 - val_loss: 0.5207 - val_accuracy: 0.8969\n",
            "Epoch 30/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2614 - accuracy: 0.9061 - val_loss: 0.5022 - val_accuracy: 0.8923\n",
            "Epoch 31/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2749 - accuracy: 0.9015 - val_loss: 0.5794 - val_accuracy: 0.8900\n",
            "Epoch 32/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2710 - accuracy: 0.8999 - val_loss: 0.6208 - val_accuracy: 0.8877\n",
            "Epoch 33/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2526 - accuracy: 0.9081 - val_loss: 0.5200 - val_accuracy: 0.8900\n",
            "Epoch 34/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2538 - accuracy: 0.9047 - val_loss: 0.5468 - val_accuracy: 0.8855\n",
            "Epoch 35/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2482 - accuracy: 0.9127 - val_loss: 0.5218 - val_accuracy: 0.8912\n",
            "Epoch 36/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2446 - accuracy: 0.9095 - val_loss: 0.5385 - val_accuracy: 0.8866\n",
            "Epoch 37/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2359 - accuracy: 0.9110 - val_loss: 0.5688 - val_accuracy: 0.8923\n",
            "Epoch 38/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2178 - accuracy: 0.9204 - val_loss: 0.5544 - val_accuracy: 0.8981\n",
            "Epoch 39/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2294 - accuracy: 0.9132 - val_loss: 0.6760 - val_accuracy: 0.8900\n",
            "Epoch 40/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2436 - accuracy: 0.9148 - val_loss: 0.4976 - val_accuracy: 0.8946\n",
            "Epoch 41/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2312 - accuracy: 0.9171 - val_loss: 0.6130 - val_accuracy: 0.8843\n",
            "Epoch 42/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2098 - accuracy: 0.9254 - val_loss: 0.5725 - val_accuracy: 0.8912\n",
            "Epoch 43/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2248 - accuracy: 0.9150 - val_loss: 0.5822 - val_accuracy: 0.8935\n",
            "Epoch 44/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2072 - accuracy: 0.9225 - val_loss: 0.5554 - val_accuracy: 0.8889\n",
            "Epoch 45/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2031 - accuracy: 0.9220 - val_loss: 0.6486 - val_accuracy: 0.8843\n",
            "Epoch 46/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2298 - accuracy: 0.9150 - val_loss: 0.6024 - val_accuracy: 0.8877\n",
            "Epoch 47/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2166 - accuracy: 0.9217 - val_loss: 0.5593 - val_accuracy: 0.8958\n",
            "Epoch 48/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1916 - accuracy: 0.9311 - val_loss: 0.5902 - val_accuracy: 0.8958\n",
            "Epoch 49/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2139 - accuracy: 0.9215 - val_loss: 0.5758 - val_accuracy: 0.8866\n",
            "Epoch 50/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2009 - accuracy: 0.9253 - val_loss: 0.6556 - val_accuracy: 0.8866\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Modelin performans metriklerini, loss ve accuracy grafiklerini ekrana yazdırın\n",
        "# Plot the the training loss\n",
        "plt.plot(results.history['loss'], label='loss')\n",
        "\n",
        "# Plot the the validation loss\n",
        "plt.plot(results.history['val_loss'], label='val_loss')\n",
        "\n",
        "# Name the x and y axises\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Loss')\n",
        "# Put legend table\n",
        "plt.legend()\n",
        "\n",
        "# Show the plot\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "aldspyIDCB_G",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "outputId": "ca42c966-19c1-4215-fd5c-163d74f793f7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the the training accuracy\n",
        "plt.plot(results.history['accuracy'], label='accuracy')\n",
        "\n",
        "# Plot the the validation accuracy\n",
        "plt.plot(results.history['val_accuracy'], label='val_accuracy')\n",
        "\n",
        "# Name the x and y axises\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Accuracy')\n",
        "\n",
        "# Put legend table\n",
        "plt.legend()\n",
        "\n",
        "# Show the plot\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "3jJ1_s2VnfII",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "outputId": "221059e0-ed3f-4eee-88d4-4d9c9886e7c6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the performance\n",
        "model.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "id": "zFj4fdVynlP7",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "18106dde-90f8-4fd0-9532-2a288d8c4e0d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "28/28 [==============================] - 1s 18ms/step - loss: 0.7118 - accuracy: 0.8833\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.7117721438407898, 0.8832951784133911]"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_result = model.predict(X_test[69].reshape(1,128,128,1))"
      ],
      "metadata": {
        "id": "4HcXQl2wq7fb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Find the predicted class\n",
        "predicted_class = prediction_result.argmax()\n",
        "# Find the prediction probability\n",
        "predicted_probability = prediction_result.max()"
      ],
      "metadata": {
        "id": "cz50naxdntpA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Print the results\n",
        "print(f'This image belongs to class {predicted_class} with {predicted_probability} probability % ')\n",
        "print(f'The real class of this image is class {y_test[69]}')"
      ],
      "metadata": {
        "id": "cxZMzMa-nyn8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b1ac1b55-a316-4b1d-b436-67c5a07b35b0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "This image belongs to class 4 with 0.9948891401290894 probability % \n",
            "The real class of this image is class 4\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Aldığınız sonuçlara göre hyperparameter optimization yapın\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "model2 = tf.keras.Sequential()\n",
        "\n",
        "model2.add(tf.keras.layers.Conv2D(64,\n",
        "                                 kernel_size = (3,3),\n",
        "                                 strides = (1,1),\n",
        "                                 padding = 'same',\n",
        "                                 activation = 'relu',\n",
        "                                 input_shape = (128,128,1)))\n",
        "                                  \n",
        "\n",
        "model2.add(tf.keras.layers.MaxPool2D(2,2))\n",
        "\n",
        "model2.add(tf.keras.layers.Conv2D(64,\n",
        "                                 kernel_size = (3,3),\n",
        "                                 strides = (1,1),\n",
        "                                 padding = 'same',\n",
        "                                 activation = 'relu',\n",
        "                                 ))\n",
        "model2.add(tf.keras.layers.MaxPool2D(2,2))\n",
        "model2.add(tf.keras.layers.Conv2D(64,\n",
        "                                 kernel_size = (3,3),\n",
        "                                 strides = (1,1),\n",
        "                                 padding = 'same',\n",
        "                                 activation = 'relu',\n",
        "                                 ))\n",
        "\n",
        "model2.add(tf.keras.layers.Flatten())\n",
        "\n",
        "model2.add(tf.keras.layers.Dense(128, activation= 'relu'))\n",
        "model2.add(tf.keras.layers.Dropout(0.5))\n",
        "model2.add(tf.keras.layers.Dense(64, activation= 'relu'))\n",
        "model2.add(tf.keras.layers.Dropout(0.5))\n",
        "\n",
        "\n",
        "model2.add(tf.keras.layers.Dense(64, activation= 'relu'))\n",
        "model2.add(tf.keras.layers.Dropout(0.5))\n",
        "\n",
        "\n",
        "model2.add(tf.keras.layers.Dense(10,activation='softmax'))\n",
        "\n",
        "\n",
        "new_optimizer = Adam(learning_rate = 0.001)\n",
        "model2.compile(optimizer= new_optimizer,\n",
        "              loss='sparse_categorical_crossentropy',\n",
        "              metrics=['accuracy']\n",
        "              )"
      ],
      "metadata": {
        "id": "9Dyz3u8aCFzC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "results2= model2.fit(X_train, y_train,\n",
        "                   batch_size=128,\n",
        "                   epochs=50,\n",
        "                   validation_data= (X_val, y_val)\n",
        "                   )"
      ],
      "metadata": {
        "id": "ItPjJk4wx0kz",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "68cb1405-f5a6-4624-8916-01c6769ad60d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "55/55 [==============================] - 6s 97ms/step - loss: 2.2104 - accuracy: 0.1742 - val_loss: 1.9184 - val_accuracy: 0.4227\n",
            "Epoch 2/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 1.9146 - accuracy: 0.3011 - val_loss: 1.5200 - val_accuracy: 0.5281\n",
            "Epoch 3/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 1.6478 - accuracy: 0.4034 - val_loss: 1.1718 - val_accuracy: 0.5670\n",
            "Epoch 4/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 1.4212 - accuracy: 0.4845 - val_loss: 0.9778 - val_accuracy: 0.6621\n",
            "Epoch 5/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 1.2427 - accuracy: 0.5632 - val_loss: 0.8205 - val_accuracy: 0.7354\n",
            "Epoch 6/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 1.0838 - accuracy: 0.6202 - val_loss: 0.7467 - val_accuracy: 0.7640\n",
            "Epoch 7/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 0.9425 - accuracy: 0.6677 - val_loss: 0.6876 - val_accuracy: 0.7995\n",
            "Epoch 8/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 0.8339 - accuracy: 0.7117 - val_loss: 0.5980 - val_accuracy: 0.8385\n",
            "Epoch 9/50\n",
            "55/55 [==============================] - 5s 96ms/step - loss: 0.7493 - accuracy: 0.7422 - val_loss: 0.5817 - val_accuracy: 0.8385\n",
            "Epoch 10/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.6954 - accuracy: 0.7658 - val_loss: 0.5765 - val_accuracy: 0.8259\n",
            "Epoch 11/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.6233 - accuracy: 0.7941 - val_loss: 0.5363 - val_accuracy: 0.8522\n",
            "Epoch 12/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.5680 - accuracy: 0.8143 - val_loss: 0.5308 - val_accuracy: 0.8454\n",
            "Epoch 13/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.5114 - accuracy: 0.8359 - val_loss: 0.4782 - val_accuracy: 0.8683\n",
            "Epoch 14/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.4723 - accuracy: 0.8527 - val_loss: 0.4726 - val_accuracy: 0.8648\n",
            "Epoch 15/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.4295 - accuracy: 0.8651 - val_loss: 0.4988 - val_accuracy: 0.8751\n",
            "Epoch 16/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.3977 - accuracy: 0.8762 - val_loss: 0.5072 - val_accuracy: 0.8740\n",
            "Epoch 17/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.3655 - accuracy: 0.8855 - val_loss: 0.5115 - val_accuracy: 0.8717\n",
            "Epoch 18/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.3379 - accuracy: 0.8938 - val_loss: 0.4946 - val_accuracy: 0.8717\n",
            "Epoch 19/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 0.3158 - accuracy: 0.9022 - val_loss: 0.4875 - val_accuracy: 0.8900\n",
            "Epoch 20/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 0.3305 - accuracy: 0.8998 - val_loss: 0.4898 - val_accuracy: 0.8751\n",
            "Epoch 21/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 0.2982 - accuracy: 0.9068 - val_loss: 0.4750 - val_accuracy: 0.8820\n",
            "Epoch 22/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2728 - accuracy: 0.9164 - val_loss: 0.5068 - val_accuracy: 0.8855\n",
            "Epoch 23/50\n",
            "55/55 [==============================] - 5s 94ms/step - loss: 0.2377 - accuracy: 0.9256 - val_loss: 0.5337 - val_accuracy: 0.8958\n",
            "Epoch 24/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2677 - accuracy: 0.9213 - val_loss: 0.4820 - val_accuracy: 0.8923\n",
            "Epoch 25/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2236 - accuracy: 0.9320 - val_loss: 0.5292 - val_accuracy: 0.8912\n",
            "Epoch 26/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2207 - accuracy: 0.9326 - val_loss: 0.5594 - val_accuracy: 0.8958\n",
            "Epoch 27/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2359 - accuracy: 0.9281 - val_loss: 0.5298 - val_accuracy: 0.8946\n",
            "Epoch 28/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2351 - accuracy: 0.9310 - val_loss: 0.5514 - val_accuracy: 0.8877\n",
            "Epoch 29/50\n",
            "55/55 [==============================] - 5s 92ms/step - loss: 0.2056 - accuracy: 0.9359 - val_loss: 0.6634 - val_accuracy: 0.8797\n",
            "Epoch 30/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2223 - accuracy: 0.9329 - val_loss: 0.6356 - val_accuracy: 0.8706\n",
            "Epoch 31/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2202 - accuracy: 0.9350 - val_loss: 0.4839 - val_accuracy: 0.9038\n",
            "Epoch 32/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1865 - accuracy: 0.9435 - val_loss: 0.5532 - val_accuracy: 0.8912\n",
            "Epoch 33/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1867 - accuracy: 0.9435 - val_loss: 0.6178 - val_accuracy: 0.8912\n",
            "Epoch 34/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2002 - accuracy: 0.9394 - val_loss: 0.6658 - val_accuracy: 0.8855\n",
            "Epoch 35/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2023 - accuracy: 0.9406 - val_loss: 0.6060 - val_accuracy: 0.8912\n",
            "Epoch 36/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.2058 - accuracy: 0.9393 - val_loss: 0.5572 - val_accuracy: 0.8935\n",
            "Epoch 37/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1744 - accuracy: 0.9503 - val_loss: 0.5579 - val_accuracy: 0.9026\n",
            "Epoch 38/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1665 - accuracy: 0.9473 - val_loss: 0.6385 - val_accuracy: 0.8923\n",
            "Epoch 39/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1658 - accuracy: 0.9495 - val_loss: 0.6332 - val_accuracy: 0.8889\n",
            "Epoch 40/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1662 - accuracy: 0.9529 - val_loss: 0.5861 - val_accuracy: 0.8981\n",
            "Epoch 41/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1692 - accuracy: 0.9493 - val_loss: 0.7119 - val_accuracy: 0.8912\n",
            "Epoch 42/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1444 - accuracy: 0.9572 - val_loss: 0.6212 - val_accuracy: 0.8981\n",
            "Epoch 43/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1616 - accuracy: 0.9513 - val_loss: 0.5881 - val_accuracy: 0.8946\n",
            "Epoch 44/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1511 - accuracy: 0.9532 - val_loss: 0.7555 - val_accuracy: 0.8786\n",
            "Epoch 45/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1664 - accuracy: 0.9520 - val_loss: 0.7113 - val_accuracy: 0.8786\n",
            "Epoch 46/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1655 - accuracy: 0.9535 - val_loss: 0.5740 - val_accuracy: 0.8900\n",
            "Epoch 47/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1551 - accuracy: 0.9543 - val_loss: 0.5999 - val_accuracy: 0.8935\n",
            "Epoch 48/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1580 - accuracy: 0.9545 - val_loss: 0.7192 - val_accuracy: 0.8866\n",
            "Epoch 49/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1682 - accuracy: 0.9487 - val_loss: 0.7065 - val_accuracy: 0.8820\n",
            "Epoch 50/50\n",
            "55/55 [==============================] - 5s 93ms/step - loss: 0.1428 - accuracy: 0.9578 - val_loss: 0.6718 - val_accuracy: 0.8843\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(results2.history['loss'], label='loss')\n",
        "\n",
        "# Plot the the validation loss\n",
        "plt.plot(results2.history['val_loss'], label='val_loss')\n",
        "\n",
        "# Name the x and y axises\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Loss')\n",
        "# Put legend table\n",
        "plt.legend()\n",
        "\n",
        "# Show the plot\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "lIFDRADzyCMM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "outputId": "675734a9-f56f-4a5f-8f93-b4b44c815848"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the the training accuracy\n",
        "plt.plot(results2.history['accuracy'], label='accuracy')\n",
        "\n",
        "# Plot the the validation accuracy\n",
        "plt.plot(results2.history['val_accuracy'], label='val_accuracy')\n",
        "\n",
        "# Name the x and y axises\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Accuracy')\n",
        "\n",
        "# Put legend table\n",
        "plt.legend()\n",
        "\n",
        "# Show the plot\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "DC23E1aUyCTA",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "outputId": "0749d96f-0044-46a2-8525-4e600034914a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model2.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "id": "sjHj8_IFzVtk",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "84385443-fcf9-4a87-ceed-531e2f1f61a4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "28/28 [==============================] - 0s 14ms/step - loss: 1.1065 - accuracy: 0.7311\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.10652756690979, 0.731121301651001]"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_result2 = model2.predict(X_test[69].reshape(1,128,128,1))"
      ],
      "metadata": {
        "id": "BEvcv0OXzVvk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "predicted_class2 = prediction_result2.argmax()\n",
        "predicted_probability2 = prediction_result2.max()"
      ],
      "metadata": {
        "id": "nlaBw9xozVyC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(f'This image belongs to class {predicted_class2} with {predicted_probability2} probability % ')\n",
        "print(f'The real class of this image is class {y_test[69]}')"
      ],
      "metadata": {
        "id": "Co91LbFIzV4o",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "943da4dd-044f-46cd-b7ab-fe1d01fd1dd0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "This image belongs to class 4 with 0.7342289090156555 probability % \n",
            "The real class of this image is class 4\n"
          ]
        }
      ]
    }
  ]
}