model_2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy of model_2.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.5"
    },
    "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/JPedro2/ImageClassificationWorkshop/blob/master/model_2.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wzNCXgzWH267",
        "colab_type": "code",
        "outputId": "e84297c8-5ee7-4a15-ecc0-c06b465415e8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from __future__ import absolute_import, division, print_function\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "\n",
        "# Helper libraries\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "print(tf.__version__)"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1.13.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r3sDTY-CH26_",
        "colab_type": "code",
        "outputId": "30ddf2e6-4274-4d65-824b-eb9754bee6d9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "fashion_mnist = keras.datasets.fashion_mnist\n",
        "\n",
        "(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()\n",
        "\n",
        "class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat', \n",
        "               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
            "32768/29515 [=================================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
            "26427392/26421880 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
            "8192/5148 [===============================================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
            "4423680/4422102 [==============================] - 0s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CwIdGA1XH27B",
        "colab_type": "code",
        "outputId": "b1152161-4cdc-4293-f2ab-cb82cf0d2574",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train_images = train_images / 255.0\n",
        "\n",
        "test_images = test_images / 255.0\n",
        "\n",
        "train_images.shape"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vMS2onp-H27D",
        "colab_type": "text"
      },
      "source": [
        "# Add depth dimension"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wOIgMN4GH27E",
        "colab_type": "code",
        "outputId": "0cc67eae-624e-4283-d983-d4b1f4cad7de",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train_images = train_images.reshape(train_images.shape[0], train_images.shape[1], train_images.shape[2], 1)\n",
        "train_labels = train_labels.reshape(train_labels.shape[0], 1)\n",
        "\n",
        "test_images = test_images.reshape(test_images.shape[0], test_images.shape[1], test_images.shape[2], 1)\n",
        "test_labels = test_labels.reshape(test_labels.shape[0], 1)\n",
        "\n",
        "train_images.shape"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28, 1)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0j0dM3HgH27H",
        "colab_type": "text"
      },
      "source": [
        "# Add Convolutional Layer and MaxPooling Layer"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pBoeGGbqH27H",
        "colab_type": "code",
        "outputId": "0879fa8e-ba2d-4e53-c501-c632f5e051ae",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "model = keras.Sequential([\n",
        "        keras.layers.Conv2D(filters = 32, kernel_size = (5,5),activation ='relu', padding='same', \n",
        "                            input_shape = (28,28,1)),\n",
        "        keras.layers.MaxPooling2D(pool_size=(2,2)),       \n",
        "\n",
        "        keras.layers.Flatten(),\n",
        "        keras.layers.Dense(512, activation = \"relu\"),\n",
        "        keras.layers.Dense(10, activation = \"softmax\")])"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RVZn7jeRH27J",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(optimizer='adam', \n",
        "              loss='sparse_categorical_crossentropy',\n",
        "              metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SeGw5HTIH27M",
        "colab_type": "code",
        "outputId": "a2937545-cb32-4bd2-b9c9-6f191b0818c0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 731
        }
      },
      "source": [
        "from time import time\n",
        "\n",
        "tensorboard = tf.keras.callbacks.TensorBoard(log_dir=\"logs/{}\".format(time()))\n",
        "\n",
        "model.fit(train_images, train_labels, epochs=20, validation_split=0.2, callbacks=[tensorboard])"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 48000 samples, validate on 12000 samples\n",
            "Epoch 1/20\n",
            "48000/48000 [==============================] - 11s 231us/sample - loss: 0.3811 - acc: 0.8633 - val_loss: 0.3171 - val_acc: 0.8866\n",
            "Epoch 2/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.2559 - acc: 0.9058 - val_loss: 0.2891 - val_acc: 0.8953\n",
            "Epoch 3/20\n",
            "48000/48000 [==============================] - 7s 155us/sample - loss: 0.2038 - acc: 0.9234 - val_loss: 0.2458 - val_acc: 0.9130\n",
            "Epoch 4/20\n",
            "48000/48000 [==============================] - 7s 155us/sample - loss: 0.1656 - acc: 0.9374 - val_loss: 0.2509 - val_acc: 0.9151\n",
            "Epoch 5/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.1319 - acc: 0.9510 - val_loss: 0.2854 - val_acc: 0.9058\n",
            "Epoch 6/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.1067 - acc: 0.9605 - val_loss: 0.2741 - val_acc: 0.9141\n",
            "Epoch 7/20\n",
            "48000/48000 [==============================] - 7s 155us/sample - loss: 0.0853 - acc: 0.9690 - val_loss: 0.2947 - val_acc: 0.9170\n",
            "Epoch 8/20\n",
            "48000/48000 [==============================] - 8s 156us/sample - loss: 0.0644 - acc: 0.9769 - val_loss: 0.3130 - val_acc: 0.9147\n",
            "Epoch 9/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.0544 - acc: 0.9805 - val_loss: 0.3534 - val_acc: 0.9162\n",
            "Epoch 10/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.0450 - acc: 0.9844 - val_loss: 0.3693 - val_acc: 0.9144\n",
            "Epoch 11/20\n",
            "48000/48000 [==============================] - 7s 155us/sample - loss: 0.0362 - acc: 0.9868 - val_loss: 0.3972 - val_acc: 0.9126\n",
            "Epoch 12/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.0336 - acc: 0.9886 - val_loss: 0.4415 - val_acc: 0.9112\n",
            "Epoch 13/20\n",
            "48000/48000 [==============================] - 8s 161us/sample - loss: 0.0276 - acc: 0.9904 - val_loss: 0.4364 - val_acc: 0.9101\n",
            "Epoch 14/20\n",
            "48000/48000 [==============================] - 8s 159us/sample - loss: 0.0209 - acc: 0.9927 - val_loss: 0.5065 - val_acc: 0.9091\n",
            "Epoch 15/20\n",
            "48000/48000 [==============================] - 7s 155us/sample - loss: 0.0209 - acc: 0.9930 - val_loss: 0.5013 - val_acc: 0.9134\n",
            "Epoch 16/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.0191 - acc: 0.9935 - val_loss: 0.5660 - val_acc: 0.9062\n",
            "Epoch 17/20\n",
            "48000/48000 [==============================] - 7s 155us/sample - loss: 0.0205 - acc: 0.9932 - val_loss: 0.5246 - val_acc: 0.9132\n",
            "Epoch 18/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.0167 - acc: 0.9944 - val_loss: 0.5291 - val_acc: 0.9118\n",
            "Epoch 19/20\n",
            "48000/48000 [==============================] - 7s 156us/sample - loss: 0.0157 - acc: 0.9950 - val_loss: 0.5354 - val_acc: 0.9123\n",
            "Epoch 20/20\n",
            "48000/48000 [==============================] - 8s 157us/sample - loss: 0.0136 - acc: 0.9956 - val_loss: 0.5695 - val_acc: 0.9120\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f04d8ad00b8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z_05aMWaH27R",
        "colab_type": "code",
        "outputId": "455530ca-b66d-486c-8c48-02826c69d857",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "test_loss, test_acc = model.evaluate(test_images, test_labels)\n",
        "\n",
        "print('Test accuracy:', test_acc)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10000/10000 [==============================] - 1s 55us/sample - loss: 0.6351 - acc: 0.9065\n",
            "Test accuracy: 0.9065\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aQrgO7UbcqeJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predictions = model.predict(test_images)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oc77J529N8e1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plot_image(i, predictions_array, true_label, img):\n",
        "  predictions_array, true_label, img = predictions_array[i], true_label[i], img[i]\n",
        "  plt.grid(False)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  \n",
        "  plt.imshow(img, cmap=plt.cm.binary)\n",
        "  \n",
        "  predicted_label = np.argmax(predictions_array)\n",
        "  if predicted_label == true_label:\n",
        "    color = 'blue'\n",
        "  else:\n",
        "    color = 'red'\n",
        "  \n",
        "  plt.xlabel(\"{} {:2.0f}% ({})\".format(class_names[predicted_label],\n",
        "                                100*np.max(predictions_array),\n",
        "                                class_names[true_label]),\n",
        "                                color=color)\n",
        "\n",
        "def plot_value_array(i, predictions_array, true_label):\n",
        "  predictions_array, true_label = predictions_array[i], true_label[i]\n",
        "  plt.grid(False)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  thisplot = plt.bar(range(10), predictions_array, color=\"#777777\")\n",
        "  plt.ylim([0, 1])\n",
        "  predicted_label = np.argmax(predictions_array)\n",
        "  \n",
        "  thisplot[predicted_label].set_color('red')\n",
        "  thisplot[true_label].set_color('blue')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "K6r4OdWCedIs",
        "colab_type": "text"
      },
      "source": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hPLPV2_Ub575",
        "colab_type": "code",
        "outputId": "94aec5d5-dc4b-4a84-b924-ef31dcb0bc0d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 592
        }
      },
      "source": [
        "# Plot the first X test images, their predicted label, and the true label\n",
        "# Color correct predictions in blue, incorrect predictions in red\n",
        "# Reshape test_data (remove last column) to allow to plot the Prediction Graph\n",
        "test_images = test_images.reshape(test_images.shape[0], test_images.shape[1], test_images.shape[2])\n",
        "test_labels = test_labels.reshape(test_labels.shape[0])\n",
        "num_rows = 5\n",
        "num_cols = 3\n",
        "num_images = num_rows*num_cols\n",
        "plt.figure(figsize=(2*2*num_cols, 2*num_rows))\n",
        "for i in range(num_images):\n",
        "  plt.subplot(num_rows, 2*num_cols, 2*i+1)\n",
        "  plot_image(i, predictions, test_labels, test_images)\n",
        "  plt.subplot(num_rows, 2*num_cols, 2*i+2)\n",
        "  plot_value_array(i, predictions, test_labels)\n",
        "plt.show()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsoAAAI/CAYAAABuyf/qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 864x720 with 30 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "source": [
        "# Next Steps\n",
        "\n",
        "Adding the convolutional layers does seem to improve our model and it seems to achieve a slightly higher accuracy. However, it is clear that the `validation loss` is a bit too high, which indicates that our model seems to be overfitting as it is learning too much from the trainning data and not able to _generalise_ so well with new data. Perhaps the model is being trained for too long, so let's look at a few techniques to avoid overfitting. \n",
        "\n",
        "[Prev](https://github.com/JPedro2/ImageClassificationWorkshop/blob/master/model_1.ipynb) - [Next](https://github.com/JPedro2/ImageClassificationWorkshop/blob/master/model_3.ipynb)"
      ],
      "cell_type": "markdown",
      "metadata": {}
    }
  ]
}