Task_4_Handwritten_digit_classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Task 4 : Handwritten_digit_classification.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sCnpBb9hV-1K"
      },
      "source": [
        "\n",
        "<h1>Task 2 : Develop A Neural Network That Can Read Handwriting</h1>\n",
        "<p>Krithika Naidu | Data Science<p>\n",
        "# LGMVIP"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_2AQSXboV6GJ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1ecea155-d1ce-49af-a311-411e09dbea8b"
      },
      "source": [
        "import tensorflow as tf\n",
        "(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 0s 0us/step\n",
            "11501568/11490434 [==============================] - 0s 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0qAftEordyBD",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        },
        "outputId": "c058b5d3-67d5-4463-fc38-b879dc1980a2"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "image_index = 7777 # You may select anything up to 60,000\n",
        "print(y_train[image_index]) # The label is 8\n",
        "plt.imshow(x_train[image_index], cmap='Greys')"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "8\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f2d23a630d0>"
            ]
          },
          "metadata": {},
          "execution_count": 5
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iN8ii2eod1-d",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1a8e9ac6-8c88-4b22-acf9-25905c2d655c"
      },
      "source": [
        "x_train.shape"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hzFNgc_qd6qd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c48e7ae6-b1eb-4d49-ce2b-089ddbb08496"
      },
      "source": [
        "# Reshaping the array to 4-dims so that it works with the Keras API\n",
        "x_train = x_train.reshape(x_train.shape[0], 28, 28, 1)\n",
        "x_test = x_test.reshape(x_test.shape[0], 28, 28, 1)\n",
        "input_shape = (28, 28, 1)\n",
        "# Making sure that the values are float so that we can get decimal points after division\n",
        "x_train = x_train.astype('float32')\n",
        "x_test = x_test.astype('float32')\n",
        "# Normalizing the RGB codes by dividing it to the max RGB value.\n",
        "x_train /= 255\n",
        "x_test /= 255\n",
        "print('x_train shape:', x_train.shape)\n",
        "print('Number of images in x_train', x_train.shape[0])\n",
        "print('Number of images in x_test', x_test.shape[0])"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "x_train shape: (60000, 28, 28, 1)\n",
            "Number of images in x_train 60000\n",
            "Number of images in x_test 10000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FGLFmfCcd9Qj"
      },
      "source": [
        "# Importing the required Keras modules containing model and layers\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Conv2D, Dropout, Flatten, MaxPooling2D\n",
        "# Creating a Sequential Model and adding the layers\n",
        "model = Sequential()\n",
        "model.add(Conv2D(28, kernel_size=(3,3), input_shape=input_shape))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "model.add(Flatten()) # Flattening the 2D arrays for fully connected layers\n",
        "model.add(Dense(128, activation=tf.nn.relu))\n",
        "model.add(Dropout(0.2))\n",
        "model.add(Dense(10,activation=tf.nn.softmax))"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a76j84Dmd_nO",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2ca6a533-d533-4f17-ef5b-667706d11974"
      },
      "source": [
        "model.compile(optimizer='adam', \n",
        "              loss='sparse_categorical_crossentropy', \n",
        "              metrics=['accuracy'])\n",
        "model.fit(x=x_train,y=y_train, epochs=10)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "1875/1875 [==============================] - 34s 18ms/step - loss: 0.2130 - accuracy: 0.9366\n",
            "Epoch 2/10\n",
            "1875/1875 [==============================] - 33s 17ms/step - loss: 0.0857 - accuracy: 0.9737\n",
            "Epoch 3/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.0571 - accuracy: 0.9824\n",
            "Epoch 4/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.0447 - accuracy: 0.9856\n",
            "Epoch 5/10\n",
            "1875/1875 [==============================] - 31s 17ms/step - loss: 0.0342 - accuracy: 0.9885\n",
            "Epoch 6/10\n",
            "1875/1875 [==============================] - 31s 17ms/step - loss: 0.0295 - accuracy: 0.9898\n",
            "Epoch 7/10\n",
            "1875/1875 [==============================] - 31s 17ms/step - loss: 0.0237 - accuracy: 0.9918\n",
            "Epoch 8/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.0228 - accuracy: 0.9923\n",
            "Epoch 9/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.0205 - accuracy: 0.9932\n",
            "Epoch 10/10\n",
            "1875/1875 [==============================] - 31s 17ms/step - loss: 0.0179 - accuracy: 0.9936\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f2d233a4f10>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P3qTGC2FeDa3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "dc8fc33b-3e88-4ad8-af97-0faa31e46218"
      },
      "source": [
        "model.evaluate(x_test, y_test)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 2s 7ms/step - loss: 0.0735 - accuracy: 0.9831\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.07352874428033829, 0.9830999970436096]"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iCPC4RRYeId_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "5c0901ca-adbc-4182-a5e4-8f0a58c2ea1e"
      },
      "source": [
        "image_index = 4445 # digit 2 \n",
        "plt.imshow(x_test[image_index].reshape(28, 28),cmap='Greys')\n",
        "pred = model.predict(x_test[image_index].reshape(1, 28, 28, 1))\n",
        "print(pred.argmax())"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD4CAYAAAAq5pAIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAANcElEQVR4nO3db6hc9Z3H8c/HmCIxPkg2lySk101TzAPZuLaOslKRbspG4x9iHyiNWhSUW8RAhcqudB80iAuyrlsEl5J0GxqX/KHY1gYUVleKIQjFiWRjori6ISGJMbnZiE18oCb97oN7lGu885ubmTN/mu/7BcPMnO+ce76cm0/O3PObOT9HhACc/y4YdAMA+oOwA0kQdiAJwg4kQdiBJC7s58bmzZsXixcv7ucmgVT279+v48ePe6paV2G3faOkpyTNkPTvEfF46fWLFy9Ws9nsZpMAChqNRstax2/jbc+Q9G+SVkq6XNJq25d3+vMA9FY3f7NfI+ndiNgXEZ9I2ippVT1tAahbN2FfJOngpOeHqmVfYHvMdtN2c3x8vIvNAehGz8/GR8T6iGhERGNkZKTXmwPQQjdhPyxpdNLzr1bLAAyhbsL+mqTLbH/N9lckfU/StnraAlC3jofeIuK07TWS/lMTQ28bImJvbZ0BqFVX4+wR8YKkF2rqBUAP8XFZIAnCDiRB2IEkCDuQBGEHkiDsQBKEHUiCsANJEHYgCcIOJEHYgSQIO5AEYQeS6OulpNEbH374YcvaY489Vlz3ySefLNZvv/32Yn3nzp3F+o4dO1rWFixYUFwX9eLIDiRB2IEkCDuQBGEHkiDsQBKEHUiCsANJMM5+Hli7dm3L2lNPPVVc155ydt/PPfvss5209LmlS5e2rO3dW77y+OjoaLGOc8ORHUiCsANJEHYgCcIOJEHYgSQIO5AEYQeSYJz9PLBs2bKWtZtvvrm47vLly4v1efPmFevr1q0r1l999dWWtauvvrq47r59+4r1WbNmFev4oq7Cbnu/pJOSzkg6HRGNOpoCUL86jux/GxHHa/g5AHqIv9mBJLoNe0h60fZO22NTvcD2mO2m7eb4+HiXmwPQqW7Dfl1EfFPSSkkP2r7+7BdExPqIaEREY2RkpMvNAehUV2GPiMPV/TFJv5V0TR1NAahfx2G3fbHtSz57LGmFpD11NQagXo6Izla0l2jiaC5NnNXfHBH/VFqn0WhEs9nsaHsYTmfOnCnW77333pa1zZs3F9fdunVrsd7umvYZNRoNNZvNKS9S0PHQW0Tsk/TXHXcFoK8YegOSIOxAEoQdSIKwA0kQdiAJvuKKrsyYMaNYf+KJJ1rWNm3aVFz3mWeeKdYZejs3HNmBJAg7kARhB5Ig7EAShB1IgrADSRB2IAnG2dFTpasTXXvttcV1L7iAY1Gd2JtAEoQdSIKwA0kQdiAJwg4kQdiBJAg7kATj7Oip0vfd77zzzuK6L774Yt3tpMaRHUiCsANJEHYgCcIOJEHYgSQIO5AEYQeSYJwdA3PJJZcU6++9916x3m666HbXtM+m7ZHd9gbbx2zvmbRsru2XbL9T3c/pbZsAujWdt/G/lHTjWcsekfRyRFwm6eXqOYAh1jbsEbFd0omzFq+StLF6vFHSbTX3BaBmnZ6gmx8RR6rH70ua3+qFtsdsN203x8fHO9wcgG51fTY+IkJSFOrrI6IREY3SxQcB9FanYT9qe6EkVffH6msJQC90GvZtku6pHt8j6Xf1tAOgV9qOs9veIunbkubZPiTpJ5Iel/Qr2/dJOiDpjl42ifNTu3H0BQsWFOuMo5+btmGPiNUtSt+puRcAPcTHZYEkCDuQBGEHkiDsQBKEHUiCr7hiYE6ePFmsM2VzvdibQBKEHUiCsANJEHYgCcIOJEHYgSQIO5AE4+zoqU8++aRlbcuWLcV1ly1bVnc7qXFkB5Ig7EAShB1IgrADSRB2IAnCDiRB2IEkGGf/M3Dq1Kliffbs2X3q5NydPn26Ze3AgQPFdW+44Ya620mNIzuQBGEHkiDsQBKEHUiCsANJEHYgCcIOJME4ex9s27atWF+zZk2xfujQoWL94Ycfblk7fvx4cd0HHnigWL/qqquK9XbXdl+3bl3LWkQU1223X3Bu2h7ZbW+wfcz2nknL1to+bHtXdbupt20C6NZ03sb/UtKNUyz/aURcWd1eqLctAHVrG/aI2C7pRB96AdBD3ZygW2N7d/U2f06rF9kes9203RwfH+9icwC60WnYfybp65KulHRE0pOtXhgR6yOiERGNkZGRDjcHoFsdhT0ijkbEmYj4k6SfS7qm3rYA1K2jsNteOOnpdyXtafVaAMPB7cY6bW+R9G1J8yQdlfST6vmVkkLSfkk/iIgj7TbWaDSi2Wx21fAw+vjjj4v1Sy+9tFhvNxbejWn8fov166+/vlhv9136559/vmXtoosuKq779ttvF+ujo6PFekaNRkPNZnPKX2rbD9VExOopFv+i664A9BUflwWSIOxAEoQdSIKwA0kQdiAJvuJag48++qhYb/cx4XbDX4O0ffv2nv3su+++u1hnaK1eHNmBJAg7kARhB5Ig7EAShB1IgrADSRB2IAnG2Wswd+7cYv2uu+4q1jdv3lxnO3822k1Fffjw4WJ90aJFdbZz3uPIDiRB2IEkCDuQBGEHkiDsQBKEHUiCsANJMM7eB08//XSxfuGF5V/Dxo0bO952u0tJ91pp+1u3bi2uu3z58mL9/vvv76inrDiyA0kQdiAJwg4kQdiBJAg7kARhB5Ig7EASbadsrtP5OmVztz799NNi/eDBg8X67t27W9YeffTR4rpXXHFFsb5s2bJi/ZZbbinWd+zY0bI2NjZWXHfp0qXFemk6aElasmRJsX4+Kk3Z3PbIbnvU9u9tv2l7r+0fVsvn2n7J9jvV/Zy6GwdQn+m8jT8t6UcRcbmkv5H0oO3LJT0i6eWIuEzSy9VzAEOqbdgj4khEvF49PinpLUmLJK2S9NnnODdKuq1XTQLo3jmdoLO9WNI3JP1B0vyIOFKV3pc0v8U6Y7abtpvt5jwD0DvTDrvt2ZJ+LemhiPjj5FpMnOWb8kxfRKyPiEZENEZGRrpqFkDnphV22zM1EfRNEfGbavFR2wur+kJJx3rTIoA6tB1688R8whslnYiIhyYtf0LS/0XE47YfkTQ3Iv6+9LMYesunNKy4YsWK4rrtpou+9dZbi/XnnnuuWD8flYbepvN99m9J+r6kN2zvqpb9WNLjkn5l+z5JByTdUUezAHqjbdgjYoekKf+nkPSdetsB0Ct8XBZIgrADSRB2IAnCDiRB2IEkuJQ0emrmzJkta+2+orpy5cpi/YMPPuiop6w4sgNJEHYgCcIOJEHYgSQIO5AEYQeSIOxAEoyzY2BmzZpVrL/yyit96iQHjuxAEoQdSIKwA0kQdiAJwg4kQdiBJAg7kARhB5Ig7EAShB1IgrADSRB2IAnCDiRB2IEkCDuQRNuw2x61/Xvbb9rea/uH1fK1tg/b3lXdbup9uwA6NZ2LV5yW9KOIeN32JZJ22n6pqv00Iv6ld+0BqMt05mc/IulI9fik7bckLep1YwDqdU5/s9teLOkbkv5QLVpje7ftDbbntFhnzHbTdnN8fLyrZgF0btphtz1b0q8lPRQRf5T0M0lfl3SlJo78T061XkSsj4hGRDRGRkZqaBlAJ6YVdtszNRH0TRHxG0mKiKMRcSYi/iTp55Ku6V2bALo1nbPxlvQLSW9FxL9OWr5w0su+K2lP/e0BqMt0zsZ/S9L3Jb1he1e17MeSVtu+UlJI2i/pBz3pEEAtpnM2fockT1F6of52APQKn6ADkiDsQBKEHUiCsANJEHYgCcIOJEHYgSQIO5AEYQeSIOxAEoQdSIKwA0kQdiAJwg4k4Yjo38bscUkHJi2aJ+l43xo4N8Pa27D2JdFbp+rs7S8jYsrrv/U17F/auN2MiMbAGigY1t6GtS+J3jrVr954Gw8kQdiBJAYd9vUD3n7JsPY2rH1J9NapvvQ20L/ZAfTPoI/sAPqEsANJDCTstm+0/bbtd20/MogeWrG93/Yb1TTUzQH3ssH2Mdt7Ji2ba/sl2+9U91POsTeg3oZiGu/CNOMD3XeDnv6873+z254h6X8k/Z2kQ5Jek7Q6It7sayMt2N4vqRERA/8Ahu3rJZ2S9ExE/FW17J8lnYiIx6v/KOdExD8MSW9rJZ0a9DTe1WxFCydPMy7pNkn3aoD7rtDXHerDfhvEkf0aSe9GxL6I+ETSVkmrBtDH0IuI7ZJOnLV4laSN1eONmvjH0nctehsKEXEkIl6vHp+U9Nk04wPdd4W++mIQYV8k6eCk54c0XPO9h6QXbe+0PTboZqYwPyKOVI/flzR/kM1Moe003v101jTjQ7PvOpn+vFucoPuy6yLim5JWSnqwers6lGLib7BhGjud1jTe/TLFNOOfG+S+63T6824NIuyHJY1Oev7VatlQiIjD1f0xSb/V8E1FffSzGXSr+2MD7udzwzSN91TTjGsI9t0gpz8fRNhfk3SZ7a/Z/oqk70naNoA+vsT2xdWJE9m+WNIKDd9U1Nsk3VM9vkfS7wbYyxcMyzTeraYZ14D33cCnP4+Ivt8k3aSJM/L/K+kfB9FDi76WSPrv6rZ30L1J2qKJt3WfauLcxn2S/kLSy5LekfRfkuYOUW//IekNSbs1EayFA+rtOk28Rd8taVd1u2nQ+67QV1/2Gx+XBZLgBB2QBGEHkiDsQBKEHUiCsANJEHYgCcIOJPH/xbcsmgIHqLkAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_7ba5WrHglH_"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}