CNN_for_text_recognition_performance_analysis.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyMNcZhsEh0fK5gkKXpnriZR",
      "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/Phanindhra-Reddy/Machine-Learning/blob/main/CNN_for_text_recognition_performance_analysis.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**baseline cnn model for mnist**"
      ],
      "metadata": {
        "id": "P1AoBzhr_3Jz"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "Ed0s2Szm-buM"
      },
      "outputs": [],
      "source": [
        "from numpy import mean\n",
        "from numpy import std\n",
        "from matplotlib import pyplot as plt\n",
        "from sklearn.model_selection import KFold\n",
        "from tensorflow.keras.datasets import mnist\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv2D\n",
        "from tensorflow.keras.layers import MaxPooling2D\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.layers import Flatten\n",
        "from tensorflow.keras.optimizers import SGD"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "load train and test dataset"
      ],
      "metadata": {
        "id": "yBkK-jzW__z-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def load_dataset():\n",
        "\t# load dataset\n",
        "\t(trainX, trainY), (testX, testY) = mnist.load_data()\n",
        "\t# reshape dataset to have a single channel\n",
        "\ttrainX = trainX.reshape((trainX.shape[0], 28, 28, 1))\n",
        "\ttestX = testX.reshape((testX.shape[0], 28, 28, 1))\n",
        "\t# one hot encode target values\n",
        "\ttrainY = to_categorical(trainY)\n",
        "\ttestY = to_categorical(testY)\n",
        "\treturn trainX, trainY, testX, testY"
      ],
      "metadata": {
        "id": "zIrlWmvJ-tUh"
      },
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "scale pixels"
      ],
      "metadata": {
        "id": "UnE5OlU5AFUm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def prep_pixels(train, test):\n",
        "\t# convert from integers to floats\n",
        "\ttrain_norm = train.astype('float32')\n",
        "\ttest_norm = test.astype('float32')\n",
        "\t# normalize to range 0-1\n",
        "\ttrain_norm = train_norm / 255.0\n",
        "\ttest_norm = test_norm / 255.0\n",
        "\t# return normalized images\n",
        "\treturn train_norm, test_norm"
      ],
      "metadata": {
        "id": "1I9l1SzL-z5N"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Define CNN model"
      ],
      "metadata": {
        "id": "Gguf7qADAJZz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def define_model():\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', input_shape=(28, 28, 1)))\n",
        "\tmodel.add(MaxPooling2D((2, 2)))\n",
        "\tmodel.add(Flatten())\n",
        "\tmodel.add(Dense(100, activation='relu', kernel_initializer='he_uniform'))\n",
        "\tmodel.add(Dense(10, activation='softmax'))\n",
        "\t# compile model\n",
        "\topt = SGD(learning_rate=0.01, momentum=0.9)\n",
        "\tmodel.compile(optimizer=opt, loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\treturn model"
      ],
      "metadata": {
        "id": "9AMaieNo-31V"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "evaluate a model using k-fold cross-validation"
      ],
      "metadata": {
        "id": "RNjE5uq9AMuD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def evaluate_model(dataX, dataY, n_folds=5):\n",
        "\tscores, histories = list(), list()\n",
        "\t# prepare cross validation\n",
        "\tkfold = KFold(n_folds, shuffle=True, random_state=1)\n",
        "\t# enumerate splits\n",
        "\tfor train_ix, test_ix in kfold.split(dataX):\n",
        "\t\t# define model\n",
        "\t\tmodel = define_model()\n",
        "\t\t# select rows for train and test\n",
        "\t\ttrainX, trainY, testX, testY = dataX[train_ix], dataY[train_ix], dataX[test_ix], dataY[test_ix]\n",
        "\t\t# fit model\n",
        "\t\thistory = model.fit(trainX, trainY, epochs=10, batch_size=32, validation_data=(testX, testY), verbose=0)\n",
        "\t\t# evaluate model\n",
        "\t\t_, acc = model.evaluate(testX, testY, verbose=0)\n",
        "\t\tprint('> %.3f' % (acc * 100.0))\n",
        "\t\t# stores scores\n",
        "\t\tscores.append(acc)\n",
        "\t\thistories.append(history)\n",
        "\treturn scores, histories"
      ],
      "metadata": {
        "id": "U88CSArp-7-H"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "plot diagnostic learning curves"
      ],
      "metadata": {
        "id": "Rs4F_dlxAQ_r"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def summarize_diagnostics(histories):\n",
        "\tfor i in range(len(histories)):\n",
        "\t\t# plot loss\n",
        "\t\tplt.subplot(2, 1, 1)\n",
        "\t\tplt.title('Cross Entropy Loss')\n",
        "\t\tplt.plot(histories[i].history['loss'], color='blue', label='train')\n",
        "\t\tplt.plot(histories[i].history['val_loss'], color='orange', label='test')\n",
        "\t\t# plot accuracy\n",
        "\t\tplt.subplot(2, 1, 2)\n",
        "\t\tplt.title('Classification Accuracy')\n",
        "\t\tplt.plot(histories[i].history['accuracy'], color='blue', label='train')\n",
        "\t\tplt.plot(histories[i].history['val_accuracy'], color='orange', label='test')\n",
        "\tplt.show()"
      ],
      "metadata": {
        "id": "x-X6wUdF_AO1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "summarize model performance"
      ],
      "metadata": {
        "id": "s3n5LcdZAWCW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def summarize_performance(scores):\n",
        "\t# print summary\n",
        "\tprint('Accuracy: mean=%.3f std=%.3f, n=%d' % (mean(scores)*100, std(scores)*100, len(scores)))\n",
        "\t# box and whisker plots of results\n",
        "\tplt.boxplot(scores)\n",
        "\tplt.show()"
      ],
      "metadata": {
        "id": "e6xgLweg_Drb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "run the test harness for evaluating a model"
      ],
      "metadata": {
        "id": "C9jm70hoAZ-T"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def run_test_harness():\n",
        "\t# load dataset\n",
        "\ttrainX, trainY, testX, testY = load_dataset()\n",
        "\t# prepare pixel data\n",
        "\ttrainX, testX = prep_pixels(trainX, testX)\n",
        "\t# evaluate model\n",
        "\tscores, histories = evaluate_model(trainX, trainY)\n",
        "\t# learning curves\n",
        "\tsummarize_diagnostics(histories)\n",
        "\t# summarize estimated performance\n",
        "\tsummarize_performance(scores)"
      ],
      "metadata": {
        "id": "_okf8CPQ_HOT"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "entry point, run the test harness"
      ],
      "metadata": {
        "id": "ax5C3XYeAfHR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "run_test_harness()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 757
        },
        "id": "04QPBwzO_Lxj",
        "outputId": "1a59b898-3d3c-46b3-fdc5-eb8820b5877f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11490434/11490434 [==============================] - 0s 0us/step\n",
            "> 98.642\n",
            "> 98.450\n",
            "> 98.567\n",
            "> 98.842\n",
            "> 98.808\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:4: MatplotlibDeprecationWarning: Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:9: MatplotlibDeprecationWarning: Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
            "  if __name__ == '__main__':\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"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy: mean=98.662 std=0.147, n=5\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}