Capstone_Project.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "Capstone Project.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.1"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/hiren-2911/capstone_project/blob/main/Capstone_Project.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Pd8UyMMyjwRZ"
      },
      "source": [
        "# Capstone Project\n",
        "## Image classifier for the SVHN dataset\n",
        "### Instructions\n",
        "\n",
        "In this notebook, you will create a neural network that classifies real-world images digits. You will use concepts from throughout this course in building, training, testing, validating and saving your Tensorflow classifier model.\n",
        "\n",
        "This project is peer-assessed. Within this notebook you will find instructions in each section for how to complete the project. Pay close attention to the instructions as the peer review will be carried out according to a grading rubric that checks key parts of the project instructions. Feel free to add extra cells into the notebook as required.\n",
        "\n",
        "### How to submit\n",
        "\n",
        "When you have completed the Capstone project notebook, you will submit a pdf of the notebook for peer review. First ensure that the notebook has been fully executed from beginning to end, and all of the cell outputs are visible. This is important, as the grading rubric depends on the reviewer being able to view the outputs of your notebook. Save the notebook as a pdf (File -> Download as -> PDF via LaTeX). You should then submit this pdf for review.\n",
        "\n",
        "### Let's get started!\n",
        "\n",
        "We'll start by running some imports, and loading the dataset. For this project you are free to make further imports throughout the notebook as you wish. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_kAJSW-ojwRc"
      },
      "source": [
        "import tensorflow as tf\n",
        "from scipy.io import loadmat\n",
        "import random\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LFpEENV8jwRd"
      },
      "source": [
        "![SVHN overview image](data/svhn_examples.jpg)\n",
        "For the capstone project, you will use the [SVHN dataset](http://ufldl.stanford.edu/housenumbers/). This is an  image dataset of over 600,000 digit images in all, and is a harder dataset than MNIST as the numbers appear in the context of natural scene images. SVHN is obtained from house numbers in Google Street View images. \n",
        "\n",
        "* Y. Netzer, T. Wang, A. Coates, A. Bissacco, B. Wu and A. Y. Ng. \"Reading Digits in Natural Images with Unsupervised Feature Learning\". NIPS Workshop on Deep Learning and Unsupervised Feature Learning, 2011.\n",
        "\n",
        "Your goal is to develop an end-to-end workflow for building, training, validating, evaluating and saving a neural network that classifies a real-world image into one of ten classes."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6F0xnz2tj8-J",
        "outputId": "d47b7633-a494-4a65-c494-81e2ee4edb37"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gMo3zNqhjwRe"
      },
      "source": [
        "# Run this cell to load the dataset\n",
        "\n",
        "train = loadmat('/content/drive/MyDrive/Coursera/train_32x32.mat')\n",
        "test = loadmat('/content/drive/MyDrive/Coursera/test_32x32.mat')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ED9U5J8RjwRf"
      },
      "source": [
        "Both `train` and `test` are dictionaries with keys `X` and `y` for the input images and labels respectively."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5VS4aulPjwRf"
      },
      "source": [
        "## 1. Inspect and preprocess the dataset\n",
        "* Extract the training and testing images and labels separately from the train and test dictionaries loaded for you.\n",
        "* Select a random sample of images and corresponding labels from the dataset (at least 10), and display them in a figure.\n",
        "* Convert the training and test images to grayscale by taking the average across all colour channels for each pixel. _Hint: retain the channel dimension, which will now have size 1._\n",
        "* Select a random sample of the grayscale images and corresponding labels from the dataset (at least 10), and display them in a figure."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2vPtfWyOjwRg"
      },
      "source": [
        "data_train=train['X']\n",
        "train_target=train['y']\n",
        "train_target=np.squeeze(train_target)\n",
        "train_data=np.ones((73257,32,32,3))\n",
        "for i in range(73257):\n",
        "    train_data[i,:,:,:]=data_train[:,:,:,i]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wNFLjvDejwRh"
      },
      "source": [
        "data_test=test['X']\n",
        "test_target=test['y']\n",
        "test_target=np.squeeze(test_target)\n",
        "test_data=np.ones((26032,32,32,3))\n",
        "for i in range(26032):\n",
        "    test_data[i,:,:,:]=data_test[:,:,:,i]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Emc6sGEwjwRh"
      },
      "source": [
        "#selecting random images\n",
        "def rnd_samp(train_data,train_target):\n",
        "    random.seed(0)\n",
        "    img_targt=list(zip(train_data,train_target))\n",
        "    random_sel=random.sample(img_targt,11)\n",
        "    rnd_img,rnd_target=zip(*random_sel)\n",
        "    rnd_img=np.array(rnd_img)\n",
        "    rnd_img=rnd_img/255\n",
        "    return rnd_img,rnd_target"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 471
        },
        "id": "FvtaP3_pjwRi",
        "scrolled": true,
        "outputId": "cde829d6-1d53-4c20-c81f-2105f249683f"
      },
      "source": [
        "rnd_img,rnd_target=rnd_samp(train_data,train_target)\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(1,11):\n",
        "    plt.subplot(2,5,i)\n",
        "    plt.imshow(rnd_img[i,:,:,:])\n",
        "    plt.title(rnd_target[i])\n",
        "plt.tight_layout()\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X6KUo3b8jwRj",
        "outputId": "e1693aac-4244-4a4d-b62b-09c4b762c8e5"
      },
      "source": [
        "train_gray=np.mean(train_data,axis=3)\n",
        "train_gray.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(73257, 32, 32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WVko7z6q_6X7",
        "outputId": "32917922-8ee6-4d1b-8ae9-ce1ff993ec88"
      },
      "source": [
        "test_gray=np.mean(test_data,axis=3)\n",
        "test_gray.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(26032, 32, 32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 471
        },
        "id": "Ih19s-ifjwRj",
        "outputId": "97940a2e-ce18-4970-9295-aedd2cf968a2"
      },
      "source": [
        "rnd_img,rnd_target=rnd_samp(train_gray,train_target)\n",
        "\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(1,11):\n",
        "    plt.subplot(2,5,i)\n",
        "    plt.imshow(rnd_img[i,:,:],cmap='gray')\n",
        "    plt.title(rnd_target[i])\n",
        "plt.tight_layout()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iJpNVEVB-PMs"
      },
      "source": [
        "train_data=train_data/255\n",
        "test_data=test_data/255"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ciQl5JaLjwRk"
      },
      "source": [
        "## 2. MLP neural network classifier\n",
        "* Build an MLP classifier model using the Sequential API. Your model should use only Flatten and Dense layers, with the final layer having a 10-way softmax output. \n",
        "* You should design and build the model yourself. Feel free to experiment with different MLP architectures. _Hint: to achieve a reasonable accuracy you won't need to use more than 4 or 5 layers._\n",
        "* Print out the model summary (using the summary() method)\n",
        "* Compile and train the model (we recommend a maximum of 30 epochs), making use of both training and validation sets during the training run. \n",
        "* Your model should track at least one appropriate metric, and use at least two callbacks during training, one of which should be a ModelCheckpoint callback.\n",
        "* As a guide, you should aim to achieve a final categorical cross entropy training loss of less than 1.0 (the validation loss might be higher).\n",
        "* Plot the learning curves for loss vs epoch and accuracy vs epoch for both training and validation sets.\n",
        "* Compute and display the loss and accuracy of the trained model on the test set."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b1Cw3FmkjwRl"
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv2D, MaxPool2D, BatchNormalization, Flatten, Dense, Dropout,Softmax "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fap2ZoXL-WnT"
      },
      "source": [
        "from keras.utils import to_categorical\n",
        "train_target_cat = to_categorical(train_target)\n",
        "test_target_cat = to_categorical(test_target)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iQRxz3UdALTM",
        "outputId": "7a44ba6f-a948-46d9-cc2c-0ec92d47dd52"
      },
      "source": [
        "image_size =train_gray.shape[1]\n",
        "input_size = image_size * image_size\n",
        "x_train = np.reshape(train_gray, [-1, input_size])\n",
        "x_train = x_train.astype('float32') / 255\n",
        "x_test = np.reshape(test_gray, [-1, input_size])\n",
        "x_test = x_test.astype('float32') / 255\n",
        "x_train.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(73257, 1024)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DFL5_7OvjwRl"
      },
      "source": [
        "def MLP(input_size):\n",
        "  model=Sequential([Dense(256, input_dim=input_size, activation='relu',kernel_initializer= tf.keras.initializers.HeUniform()),\n",
        "                    BatchNormalization(),\n",
        "                    Dense(128, activation='relu',kernel_initializer= tf.keras.initializers.HeUniform()),\n",
        "                    BatchNormalization(),\n",
        "                    Dense(128, activation='relu',kernel_initializer= tf.keras.initializers.HeUniform()),\n",
        "                    BatchNormalization(),\n",
        "                    Dropout(0.3),\n",
        "                    Dense(128, activation='relu',kernel_initializer= tf.keras.initializers.HeUniform()),\n",
        "                    BatchNormalization(),\n",
        "                    Dropout(0.3),\n",
        "                    Dense(64, activation='relu',kernel_initializer= tf.keras.initializers.HeUniform()),\n",
        "                    Dense(11, activation='softmax')])\n",
        "  model.compile(loss='categorical_crossentropy',optimizer=tf.keras.optimizers.Adam(0.01),metrics=['accuracy'])\n",
        "  return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KKr5-PfhjwRl"
      },
      "source": [
        "model=MLP(1024)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t6ei7aAZwaoN"
      },
      "source": [
        "from tensorflow.keras.callbacks import ModelCheckpoint,EarlyStopping\n",
        "Er=EarlyStopping(monitor='val_accuracy',patience=10,min_delta=0.001,mode='max')\n",
        "best_check=ModelCheckpoint('/content/drive/MyDrive/Model_weights/best_MLP',monitor='val_accuracy',save_best_only=True)\n",
        "lr=tf.keras.callbacks.ReduceLROnPlateau(monitor=\"loss\",factor=(0.4), verbose=1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FHBnJ7l3jwRl",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a428f89f-e4a3-4863-ca5f-195f02d598ad"
      },
      "source": [
        "history=model.fit(x_train,train_target_cat,validation_data=(x_test,test_target_cat),epochs=30,batch_size=64,callbacks=[best_check,lr])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/30\n",
            "1145/1145 [==============================] - 13s 8ms/step - loss: 2.1348 - accuracy: 0.2552 - val_loss: 1.9758 - val_accuracy: 0.3857\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 2/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 1.4305 - accuracy: 0.5169 - val_loss: 2.3502 - val_accuracy: 0.3126\n",
            "Epoch 3/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 1.2607 - accuracy: 0.5967 - val_loss: 1.4618 - val_accuracy: 0.5678\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 4/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 1.1569 - accuracy: 0.6416 - val_loss: 1.3525 - val_accuracy: 0.5857\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 5/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 1.1196 - accuracy: 0.6549 - val_loss: 1.3515 - val_accuracy: 0.6036\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 6/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 1.0698 - accuracy: 0.6718 - val_loss: 1.3535 - val_accuracy: 0.5804\n",
            "Epoch 7/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 1.0209 - accuracy: 0.6870 - val_loss: 1.3248 - val_accuracy: 0.5895\n",
            "Epoch 8/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.9972 - accuracy: 0.6953 - val_loss: 1.1620 - val_accuracy: 0.6311\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 9/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.9699 - accuracy: 0.7057 - val_loss: 1.0795 - val_accuracy: 0.6683\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 10/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.9573 - accuracy: 0.7068 - val_loss: 1.2097 - val_accuracy: 0.6238\n",
            "Epoch 11/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.9440 - accuracy: 0.7121 - val_loss: 1.0952 - val_accuracy: 0.6752\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 12/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.9233 - accuracy: 0.7189 - val_loss: 1.0434 - val_accuracy: 0.6871\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 13/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.9232 - accuracy: 0.7197 - val_loss: 1.1930 - val_accuracy: 0.6290\n",
            "Epoch 14/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.9119 - accuracy: 0.7223 - val_loss: 0.9890 - val_accuracy: 0.7022\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 15/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8972 - accuracy: 0.7282 - val_loss: 1.0527 - val_accuracy: 0.6782\n",
            "Epoch 16/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8994 - accuracy: 0.7274 - val_loss: 1.0178 - val_accuracy: 0.6902\n",
            "Epoch 17/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8955 - accuracy: 0.7313 - val_loss: 1.1092 - val_accuracy: 0.6564\n",
            "Epoch 18/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8863 - accuracy: 0.7317 - val_loss: 0.9431 - val_accuracy: 0.7117\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 19/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8918 - accuracy: 0.7304 - val_loss: 0.9888 - val_accuracy: 0.7008\n",
            "Epoch 20/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8763 - accuracy: 0.7350 - val_loss: 1.1017 - val_accuracy: 0.6665\n",
            "Epoch 21/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8865 - accuracy: 0.7301 - val_loss: 0.9659 - val_accuracy: 0.7015\n",
            "Epoch 22/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8617 - accuracy: 0.7382 - val_loss: 1.0010 - val_accuracy: 0.6926\n",
            "Epoch 23/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8784 - accuracy: 0.7356 - val_loss: 1.0186 - val_accuracy: 0.6893\n",
            "Epoch 24/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8702 - accuracy: 0.7365 - val_loss: 0.8945 - val_accuracy: 0.7251\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 25/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8659 - accuracy: 0.7379 - val_loss: 0.9391 - val_accuracy: 0.7104\n",
            "Epoch 26/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8680 - accuracy: 0.7380 - val_loss: 1.0060 - val_accuracy: 0.6965\n",
            "Epoch 27/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8589 - accuracy: 0.7392 - val_loss: 0.9100 - val_accuracy: 0.7238\n",
            "Epoch 28/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8581 - accuracy: 0.7408 - val_loss: 0.8747 - val_accuracy: 0.7363\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_MLP/assets\n",
            "Epoch 29/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8530 - accuracy: 0.7411 - val_loss: 0.9891 - val_accuracy: 0.6983\n",
            "Epoch 30/30\n",
            "1145/1145 [==============================] - 9s 8ms/step - loss: 0.8614 - accuracy: 0.7419 - val_loss: 1.0306 - val_accuracy: 0.6841\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jggiiun7jwRq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 284
        },
        "outputId": "ed16d714-eacc-44b6-fb30-4610d1916283"
      },
      "source": [
        "loss=history.history['loss']\n",
        "val_loss=history.history['val_loss']\n",
        "epochs=np.arange(0,30)\n",
        "plt.plot(epochs,loss)\n",
        "plt.plot(epochs,val_loss)\n",
        "plt.legend(['loss','val_loss'])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f7071e56f50>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iUYteW1ojwRq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "7d1a9b5d-d119-4ee9-d4f8-569918ac2cd4"
      },
      "source": [
        "accuracy=history.history['accuracy']\n",
        "val_accuracy=history.history['val_accuracy']\n",
        "epochs=np.arange(0,30)\n",
        "plt.plot(epochs,accuracy)\n",
        "plt.plot(epochs,val_accuracy)\n",
        "plt.legend(['accuracy','val_accuracy'])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f7071fb8450>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2U6-IFrpjwRq"
      },
      "source": [
        "## 3. CNN neural network classifier\n",
        "* Build a CNN classifier model using the Sequential API. Your model should use the Conv2D, MaxPool2D, BatchNormalization, Flatten, Dense and Dropout layers. The final layer should again have a 10-way softmax output. \n",
        "* You should design and build the model yourself. Feel free to experiment with different CNN architectures. _Hint: to achieve a reasonable accuracy you won't need to use more than 2 or 3 convolutional layers and 2 fully connected layers.)_\n",
        "* The CNN model should use fewer trainable parameters than your MLP model.\n",
        "* Compile and train the model (we recommend a maximum of 30 epochs), making use of both training and validation sets during the training run.\n",
        "* Your model should track at least one appropriate metric, and use at least two callbacks during training, one of which should be a ModelCheckpoint callback.\n",
        "* You should aim to beat the MLP model performance with fewer parameters!\n",
        "* Plot the learning curves for loss vs epoch and accuracy vs epoch for both training and validation sets.\n",
        "* Compute and display the loss and accuracy of the trained model on the test set."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D1T7Y_GJjwRr"
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv2D, MaxPool2D, BatchNormalization, Flatten, Dense, Dropout,Softmax "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nV-xTemfjwRr"
      },
      "source": [
        "def get_model(input_shape):\n",
        "    model=Sequential([Conv2D(64,(3,3),kernel_initializer= tf.keras.initializers.HeUniform(),activation='relu',input_shape=input_shape,bias_initializer='ones',padding='same'),\n",
        "                      MaxPool2D(2),\n",
        "                      Conv2D(64,(3,3),kernel_initializer=tf.keras.initializers.HeUniform(),activation='relu',bias_initializer='ones',padding='same'),\n",
        "                      MaxPool2D(2),\n",
        "                      Conv2D(64,(3,3),kernel_initializer=tf.keras.initializers.HeUniform(),activation='relu',bias_initializer='ones',padding='same'),\n",
        "                      MaxPool2D(2),\n",
        "                      Flatten(),\n",
        "                      Dense(256,activation='relu',kernel_initializer=tf.keras.initializers.HeUniform()),\n",
        "                      BatchNormalization(),\n",
        "                      Dropout(0.4),\n",
        "                      Dense(11,activation='softmax')])\n",
        "    model.compile(optimizer=tf.keras.optimizers.Adam(lr=0.01),loss='sparse_categorical_crossentropy',metrics=['accuracy'])\n",
        "    return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IqNcCPImjwRr"
      },
      "source": [
        "model_conv=get_model(train_data[0].shape)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HPkpXTwRjwRs",
        "outputId": "e6b1086d-591b-458c-e248-a1efd47d5cca"
      },
      "source": [
        "model_conv.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_3 (Conv2D)            (None, 32, 32, 64)        1792      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 16, 16, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 16, 16, 64)        36928     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 8, 8, 64)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 8, 8, 64)          36928     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 4, 4, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 1024)              0         \n",
            "_________________________________________________________________\n",
            "dense_9 (Dense)              (None, 256)               262400    \n",
            "_________________________________________________________________\n",
            "batch_normalization_5 (Batch (None, 256)               1024      \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "dense_10 (Dense)             (None, 11)                2827      \n",
            "=================================================================\n",
            "Total params: 341,899\n",
            "Trainable params: 341,387\n",
            "Non-trainable params: 512\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9xD8IgMY54At"
      },
      "source": [
        "#callbacks\n",
        "from tensorflow.keras.callbacks import ModelCheckpoint,EarlyStopping\n",
        "Er=EarlyStopping(monitor='val_accuracy',patience=10,min_delta=0.001,mode='max')\n",
        "best_check=ModelCheckpoint('/content/drive/MyDrive/Model_weights/best_normalized_initilizer',monitor='val_accuracy',save_best_only=True)\n",
        "lr=tf.keras.callbacks.ReduceLROnPlateau(monitor=\"loss\",factor=0.2, verbose=1,patience=10)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aOcpabJGjwRs",
        "outputId": "0a74e4e5-5b99-4227-f4f0-b7826eb7a817"
      },
      "source": [
        "history=model_conv.fit(train_data,train_target,epochs=30,batch_size=32,verbose=1,validation_data=(test_data,test_target),callbacks=[best_check,lr])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/30\n",
            "2290/2290 [==============================] - 24s 10ms/step - loss: 2.0125 - accuracy: 0.2998 - val_loss: 1.5377 - val_accuracy: 0.5181\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 2/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 1.2530 - accuracy: 0.5653 - val_loss: 1.6440 - val_accuracy: 0.6253\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 3/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.9518 - accuracy: 0.6902 - val_loss: 0.7179 - val_accuracy: 0.8045\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 4/30\n",
            "2290/2290 [==============================] - 24s 10ms/step - loss: 0.7725 - accuracy: 0.7581 - val_loss: 4.7582 - val_accuracy: 0.7392\n",
            "Epoch 5/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.9232 - accuracy: 0.7027 - val_loss: 1.1087 - val_accuracy: 0.6545\n",
            "Epoch 6/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.8824 - accuracy: 0.7196 - val_loss: 1.3749 - val_accuracy: 0.8227\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 7/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.6896 - accuracy: 0.7889 - val_loss: 0.6128 - val_accuracy: 0.8243\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 8/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.6015 - accuracy: 0.8185 - val_loss: 1.0518 - val_accuracy: 0.8296\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 9/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.5544 - accuracy: 0.8313 - val_loss: 0.4690 - val_accuracy: 0.8625\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 10/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.5363 - accuracy: 0.8382 - val_loss: 0.7427 - val_accuracy: 0.8521\n",
            "Epoch 11/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.5113 - accuracy: 0.8483 - val_loss: 0.6919 - val_accuracy: 0.8653\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 12/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.4924 - accuracy: 0.8541 - val_loss: 1.0106 - val_accuracy: 0.8528\n",
            "Epoch 13/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.4789 - accuracy: 0.8572 - val_loss: 0.6142 - val_accuracy: 0.8699\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 14/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.4605 - accuracy: 0.8625 - val_loss: 0.6115 - val_accuracy: 0.8710\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 15/30\n",
            "2290/2290 [==============================] - 22s 10ms/step - loss: 0.4505 - accuracy: 0.8653 - val_loss: 0.5354 - val_accuracy: 0.8624\n",
            "Epoch 16/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.4391 - accuracy: 0.8669 - val_loss: 3.2339 - val_accuracy: 0.8793\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 17/30\n",
            "2290/2290 [==============================] - 22s 10ms/step - loss: 0.4293 - accuracy: 0.8705 - val_loss: 0.6472 - val_accuracy: 0.8782\n",
            "Epoch 18/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.4235 - accuracy: 0.8735 - val_loss: 0.4270 - val_accuracy: 0.8810\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 19/30\n",
            "2290/2290 [==============================] - 22s 10ms/step - loss: 0.4064 - accuracy: 0.8786 - val_loss: 0.5026 - val_accuracy: 0.8812\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 20/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.4105 - accuracy: 0.8773 - val_loss: 0.7644 - val_accuracy: 0.8835\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 21/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.4028 - accuracy: 0.8803 - val_loss: 0.4938 - val_accuracy: 0.8892\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 22/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.3859 - accuracy: 0.8851 - val_loss: 0.7568 - val_accuracy: 0.8835\n",
            "Epoch 23/30\n",
            "2290/2290 [==============================] - 22s 10ms/step - loss: 0.3927 - accuracy: 0.8823 - val_loss: 0.7161 - val_accuracy: 0.8801\n",
            "Epoch 24/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.3846 - accuracy: 0.8866 - val_loss: 0.4432 - val_accuracy: 0.8873\n",
            "Epoch 25/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.3909 - accuracy: 0.8850 - val_loss: 0.3721 - val_accuracy: 0.8901\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 26/30\n",
            "2290/2290 [==============================] - 24s 10ms/step - loss: 0.3905 - accuracy: 0.8851 - val_loss: 0.3791 - val_accuracy: 0.8883\n",
            "Epoch 27/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.3776 - accuracy: 0.8875 - val_loss: 0.8286 - val_accuracy: 0.8898\n",
            "Epoch 28/30\n",
            "2290/2290 [==============================] - 24s 10ms/step - loss: 0.3738 - accuracy: 0.8883 - val_loss: 0.4159 - val_accuracy: 0.8825\n",
            "Epoch 29/30\n",
            "2290/2290 [==============================] - 23s 10ms/step - loss: 0.3668 - accuracy: 0.8913 - val_loss: 1.2915 - val_accuracy: 0.8904\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n",
            "Epoch 30/30\n",
            "2290/2290 [==============================] - 24s 10ms/step - loss: 0.3570 - accuracy: 0.8916 - val_loss: 0.3914 - val_accuracy: 0.8908\n",
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/Model_weights/best_normalized_initilizer/assets\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "di6c3MdWjwRt",
        "outputId": "9eaa9c36-4466-45ab-cfc6-6b37f6ca2e2c"
      },
      "source": [
        " loss=history.history['loss']\n",
        "val_loss=history.history['val_loss']\n",
        "epochs=np.arange(0,30)\n",
        "plt.plot(epochs,loss)\n",
        "plt.plot(epochs,val_loss)\n",
        "plt.legend(['loss','val_loss'])\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f7024f53b10>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "iNAouj6pOp7I",
        "outputId": "87ac402d-ee35-4d4d-d48b-6f0e1404581c"
      },
      "source": [
        "accuracy=history.history['accuracy']\n",
        "val_accuracy=history.history['val_accuracy']\n",
        "epochs=np.arange(0,30)\n",
        "plt.plot(epochs,accuracy)\n",
        "plt.plot(epochs,val_accuracy)\n",
        "plt.legend(['accuracy','val_accuracy'])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f707addb4d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 40
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xPBSk7mMjwRt"
      },
      "source": [
        "## 4. Get model predictions\n",
        "* Load the best weights for the MLP and CNN models that you saved during the training run.\n",
        "* Randomly select 5 images and corresponding labels from the test set and display the images with their labels.\n",
        "* Alongside the image and label, show each model’s predictive distribution as a bar chart, and the final model prediction given by the label with maximum probability."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V4dKHzD_jwRu"
      },
      "source": [
        "model_pred=tf.keras.models.load_model('/content/drive/MyDrive/Model_weights/best_normalized_initilizer')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tPE3BOSZjwRu",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cad8703b-f6c4-4032-a87f-7d4f8553da1a"
      },
      "source": [
        "conv_pred=np.argmax(model_pred.predict(test_data),-1)\n",
        "conv_pred"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([5, 2, 1, ..., 7, 6, 7])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 81
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LSgHFTovjwRu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 471
        },
        "outputId": "2c4b7972-8385-48bf-ea42-79313ca48f70"
      },
      "source": [
        "rnd_img,rnd_target=rnd_samp(test_data*255,conv_pred)\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(1,11):\n",
        "    plt.subplot(2,5,i)\n",
        "    plt.imshow(rnd_img[i,:,:,:])\n",
        "    plt.title(rnd_target[i])\n",
        "plt.tight_layout()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gO-dWnAjjwRu"
      },
      "source": [
        "model_mlp=tf.keras.models.load_model('/content/drive/MyDrive/Model_weights/best_MLP')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "knKv2z-ljwRu",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "408ad272-be0d-4bdc-a6b7-dbc9710178e2"
      },
      "source": [
        "mlp_pred=np.argmax(model_mlp.predict(x_test),-1)\n",
        "mlp_pred"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([5, 2, 1, ..., 2, 6, 7])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 80
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zKArKOyrjwRv",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 471
        },
        "outputId": "7c697e76-e6d1-4553-8eb7-e4e4d6c64294"
      },
      "source": [
        "rnd_img,rnd_target=rnd_samp(test_data*255,mlp_pred)\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(1,11):\n",
        "    plt.subplot(2,5,i)\n",
        "    plt.imshow(rnd_img[i,:,:,:])\n",
        "    plt.title(rnd_target[i])\n",
        "plt.tight_layout()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4BY_Tl8KDW02"
      },
      "source": [
        "ctr_conv={'0':0,'1':0,'2':0,'3':0,'4':0,'5':0,'6':0,'7':0,'8':0,'9':0,'10':0}\n",
        "for i in conv_pred:\n",
        "  ctr_conv[str(i)]=ctr_conv[str(i)]+1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "4bWrymtuDzJ2",
        "outputId": "d26539ce-1866-4e44-bb38-c994cd815acf"
      },
      "source": [
        "plt.bar(ctr_conv.keys(),ctr_conv.values())\n",
        "plt.title('Conv2D predictor Bar plot')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Conv2D predictor Bar plot')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 77
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YS3t65C8EGo2"
      },
      "source": [
        "ctr_mlp={'0':0,'1':0,'2':0,'3':0,'4':0,'5':0,'6':0,'7':0,'8':0,'9':0,'10':0}\n",
        "for i in mlp_pred:\n",
        "  ctr_mlp[str(i)]=ctr_mlp[str(i)]+1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "7tnUwUOFF90q",
        "outputId": "179e5b37-8cdf-4f66-a0e0-05c06b15d747"
      },
      "source": [
        "plt.bar(ctr_mlp.keys(),ctr_mlp.values())\n",
        "plt.title('MLP predictor Bar plot')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'MLP predictor Bar plot')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 75
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "id": "yWtaOaoQGX9-",
        "outputId": "8282c93d-2b10-41fa-fbc5-a60f051ea598"
      },
      "source": [
        "ctr_target={'0':0,'1':0,'2':0,'3':0,'4':0,'5':0,'6':0,'7':0,'8':0,'9':0,'10':0}\n",
        "for i in test_target:\n",
        "  ctr_target[str(i)]=ctr_target[str(i)]+1\n",
        "plt.bar(ctr_target.keys(),ctr_target.values())\n",
        "plt.title('Original data Bar plot')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Original data Bar plot')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 83
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8tq5eNvdIEDB"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}