Fashion_mnist_part2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyOzFQsHx1VB4X5yRIqMkpe8",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/tusharbansal22/Tensorflow-Notebooks/blob/main/Fashion_mnist_part2.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "u7TxckhV1wAa",
        "outputId": "1676da22-6ea2-4665-c862-736615a19382"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.9.2\n"
          ]
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras.layers import Input,Conv2D,Dense,Flatten,Dropout\n",
        "from tensorflow.keras.models import Model"
      ],
      "metadata": {
        "id": "y94zHHmJ2aVg"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "fashion_mnist= tf.keras.datasets.fashion_mnist\n",
        "\n",
        "(x_train,y_train),(x_test,y_test)=fashion_mnist.load_data()\n",
        "x_train,x_test=x_train/255.0,x_test/255.0\n",
        "print(x_train.shape)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FAOk53lS2p0p",
        "outputId": "0095b386-b25e-4691-acd3-b2c2416915a1"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(60000, 28, 28)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train=np.expand_dims(x_train,-1)\n",
        "x_test=np.expand_dims(x_test,-1)\n",
        "print(x_train.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uQgNadUNlWbS",
        "outputId": "a604f883-c91d-4e65-b632-14c563e7a105"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(60000, 28, 28, 1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "K=len(set(y_train))\n",
        "print('number of classes',K)"
      ],
      "metadata": {
        "id": "fv9pn_T_3BNE",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "aa5db725-90c3-4f90-e18e-e764a2a6d323"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "number of classes 10\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "i= Input(shape=x_train[0].shape)\n",
        "x=Conv2D(32,(3,3),strides=2,activation='relu')(i)\n",
        "x=Conv2D(64,(3,3),strides=2,activation='relu')(x)\n",
        "x=Conv2D(128,(3,3),strides=2,activation='relu')(x)\n",
        "x=Flatten()(x)\n",
        "x=Dropout(0.2)(x)\n",
        "x=Dense(512,activation='relu')(x)\n",
        "x=Dropout(0.2)(x)\n",
        "x=Dense(K,activation='softmax')(x)\n",
        "\n",
        "model=Model(i,x)"
      ],
      "metadata": {
        "id": "Qs7BlAQclI-N"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(optimizer='adam',loss='sparse_categorical_crossentropy',metrics=['accuracy'])\n",
        "r=model.fit(x_train,y_train,validation_data=(x_test,y_test),epochs=15)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NDuIpXK1mq2V",
        "outputId": "f4629357-ba18-4486-b225-6cc71de69648"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/15\n",
            "1875/1875 [==============================] - 28s 15ms/step - loss: 0.5124 - accuracy: 0.8096 - val_loss: 0.3994 - val_accuracy: 0.8516\n",
            "Epoch 2/15\n",
            "1875/1875 [==============================] - 43s 23ms/step - loss: 0.3522 - accuracy: 0.8672 - val_loss: 0.3339 - val_accuracy: 0.8763\n",
            "Epoch 3/15\n",
            "1875/1875 [==============================] - 42s 22ms/step - loss: 0.3048 - accuracy: 0.8857 - val_loss: 0.3255 - val_accuracy: 0.8804\n",
            "Epoch 4/15\n",
            "1875/1875 [==============================] - 36s 19ms/step - loss: 0.2726 - accuracy: 0.8970 - val_loss: 0.3163 - val_accuracy: 0.8842\n",
            "Epoch 5/15\n",
            "1875/1875 [==============================] - 31s 17ms/step - loss: 0.2495 - accuracy: 0.9041 - val_loss: 0.3124 - val_accuracy: 0.8914\n",
            "Epoch 6/15\n",
            "1875/1875 [==============================] - 27s 14ms/step - loss: 0.2272 - accuracy: 0.9143 - val_loss: 0.2945 - val_accuracy: 0.8998\n",
            "Epoch 7/15\n",
            "1875/1875 [==============================] - 27s 15ms/step - loss: 0.2105 - accuracy: 0.9202 - val_loss: 0.3005 - val_accuracy: 0.8958\n",
            "Epoch 8/15\n",
            "1875/1875 [==============================] - 29s 15ms/step - loss: 0.1952 - accuracy: 0.9257 - val_loss: 0.3032 - val_accuracy: 0.8977\n",
            "Epoch 9/15\n",
            "1875/1875 [==============================] - 36s 19ms/step - loss: 0.1815 - accuracy: 0.9307 - val_loss: 0.3234 - val_accuracy: 0.8922\n",
            "Epoch 10/15\n",
            "1875/1875 [==============================] - 36s 19ms/step - loss: 0.1693 - accuracy: 0.9348 - val_loss: 0.3172 - val_accuracy: 0.8993\n",
            "Epoch 11/15\n",
            "1875/1875 [==============================] - 33s 17ms/step - loss: 0.1567 - accuracy: 0.9401 - val_loss: 0.3296 - val_accuracy: 0.8943\n",
            "Epoch 12/15\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.1494 - accuracy: 0.9416 - val_loss: 0.3396 - val_accuracy: 0.9000\n",
            "Epoch 13/15\n",
            "1875/1875 [==============================] - 48s 25ms/step - loss: 0.1385 - accuracy: 0.9473 - val_loss: 0.3413 - val_accuracy: 0.9006\n",
            "Epoch 14/15\n",
            "1875/1875 [==============================] - 47s 25ms/step - loss: 0.1360 - accuracy: 0.9479 - val_loss: 0.3800 - val_accuracy: 0.8972\n",
            "Epoch 15/15\n",
            "1875/1875 [==============================] - 34s 18ms/step - loss: 0.1264 - accuracy: 0.9510 - val_loss: 0.3636 - val_accuracy: 0.8994\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.plot(r.history['loss'],label='loss')\n",
        "plt.plot(r.history['val_loss'],label='val_loss')\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "-qVwaU9hnJT0",
        "outputId": "32bd5ee8-7d55-43dc-e83e-076c3b51e5be"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f31981f8fd0>"
            ]
          },
          "metadata": {},
          "execution_count": 17
        },
        {
          "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": [
        "plt.plot(r.history['accuracy'],label='accuracy')\n",
        "plt.plot(r.history['val_accuracy'],label='val_accuracy')\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "MSQBbM61ngPY",
        "outputId": "064f3e60-8fa1-41d2-d5fd-9896e8435d3c"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f3197cdcac0>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        },
        {
          "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": [
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "import numpy as np\n",
        "import itertools\n",
        "\n",
        "def plot_confusion_matrix(cm,classes,normalize=False,title='Confusion Matrix',cmap=plt.cm.Blues):\n",
        "  if normalize:\n",
        "    cm=cm.astype('float')/cm.sum(axis=1)[:,np.newaxis]\n",
        "    print(\"Normalized confusion matrix\")\n",
        "  else:\n",
        "    print('Confusion matrix, without normalization')\n",
        "  print(cm)\n",
        "  plt.imshow(cm,interpolation='nearest',cmap=cmap)\n",
        "  plt.title(title)\n",
        "  plt.colorbar()\n",
        "  tick_marks=np.arange(len(classes))\n",
        "  plt.xticks(tick_marks,classes,rotation=45)\n",
        "  plt.yticks(tick_marks,classes)\n",
        "  fmt= '.2f' if normalize else 'd'\n",
        "  thresh= cm.max()/2.\n",
        "  for i,j in itertools.product(range(cm.shape[0]),range(cm.shape[1])):\n",
        "    plt.text(j,i,format(cm[i,j],fmt),\n",
        "             horizontalalignment='center',\n",
        "             color='white'if cm[i,j]>thresh else 'black')\n",
        "  plt.tight_layout()\n",
        "  plt.ylabel('True label')\n",
        "  plt.xlabel('Predicted label')\n",
        "  plt.show()\n",
        "\n",
        "p_test=model.predict(x_test).argmax(axis=1)\n",
        "cm=confusion_matrix(y_test,p_test)\n",
        "plot_confusion_matrix(cm,list(range(10)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 525
        },
        "id": "BpONIRHloYn7",
        "outputId": "ecab19ce-a0ac-431c-865c-841f52b18cc5"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 2s 6ms/step\n",
            "Confusion matrix, without normalization\n",
            "[[869   2  29  19   4   1  73   0   3   0]\n",
            " [  2 980   1   9   1   0   3   0   4   0]\n",
            " [ 11   0 877  13  52   0  46   0   1   0]\n",
            " [ 16  11  22 904  29   1  15   0   2   0]\n",
            " [  0   1  82  27 835   0  54   0   1   0]\n",
            " [  0   0   0   1   0 973   0  19   0   7]\n",
            " [127   2  82  28  86   1 662   0  12   0]\n",
            " [  0   0   0   0   0   7   0 975   0  18]\n",
            " [  3   1   5   1   3   3   8   3 973   0]\n",
            " [  0   0   0   0   0   4   1  48   1 946]]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "labels='''T-shirt/top\n",
        "Trouser\n",
        "Pullover\n",
        "Dress\n",
        "Coat\n",
        "Sandal\n",
        "Shirt\n",
        "Sneaker\n",
        "Bag\n",
        "Ankle Boot\n",
        "'''.split()"
      ],
      "metadata": {
        "id": "lHRWbaAToqwV"
      },
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "misclassified_idx=np.where(p_test!=y_test)[0]\n",
        "i=np.random.choice(misclassified_idx)\n",
        "plt.imshow(x_test[i].reshape(28,28),cmap='gray')\n",
        "plt.title('True label: %s  Predicted label: %s'%(labels[y_test[i]],labels[p_test[i]]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "D-sdgFqBodfH",
        "outputId": "e0fb0258-f030-4816-b72d-696827e82428"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'True label: Dress  Predicted label: Coat')"
            ]
          },
          "metadata": {},
          "execution_count": 25
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}