Day_48_Multilayer_Perceptron.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyNYewcmXuVpWDgU6dLcSHd9"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "taX0N_ichj7G"
      },
      "outputs": [],
      "source": [
        "%matplotlib inline\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "np.random.seed(0)\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.datasets import reuters\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Dropout, Activation\n",
        "from tensorflow.keras.preprocessing.text import Tokenizer"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Load the Data"
      ],
      "metadata": {
        "id": "rkzREp0ph3V1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "( x_train, y_train), (x_test,y_test) = reuters.load_data(num_words=10000,test_split=0.2)\n",
        "print(len(x_train))\n",
        "print(len(x_test))\n",
        "num_classes = np.max(y_train) +1\n",
        "print(num_classes)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GmKK0gbmh56E",
        "outputId": "4bf55e6b-8b09-4cc4-ac19-dcc37efed426"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/reuters.npz\n",
            "2113536/2110848 [==============================] - 0s 0us/step\n",
            "2121728/2110848 [==============================] - 0s 0us/step\n",
            "8982\n",
            "2246\n",
            "46\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Vectorize and One-hot Encoding"
      ],
      "metadata": {
        "id": "zJQv_I9ah3TE"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tokenizer = Tokenizer(num_words=10000)\n",
        "x_train = tokenizer.sequences_to_matrix(x_train,mode='binary')\n",
        "x_test = tokenizer.sequences_to_matrix(x_test,mode='binary')\n",
        "y_train = tf.keras.utils.to_categorical(y_train,num_classes)\n",
        "y_test = tf.keras.utils.to_categorical(y_test,num_classes)\n",
        "x_train.shape,x_test.shape\n",
        "y_train.shape, y_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "naTUZncmh_R1",
        "outputId": "30b20fb5-37b7-4e82-881a-c58224ebfe64"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((8982, 46), (2246, 46))"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Build Multilayer Perceptron Model"
      ],
      "metadata": {
        "id": "3jpUqZ94h3Qh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model = Sequential([\n",
        "    Dense(512,input_shape = (10000,)),\n",
        "    Activation('relu'),\n",
        "    Dropout(0.5),\n",
        "    Dense(num_classes),\n",
        "    Activation('softmax')\n",
        "    \n",
        "    \n",
        "])\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "L7XeAOUAiDUL",
        "outputId": "e64c3cb7-426a-4af5-f7bb-a8b64a732cf5"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense (Dense)               (None, 512)               5120512   \n",
            "                                                                 \n",
            " activation (Activation)     (None, 512)               0         \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 512)               0         \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 46)                23598     \n",
            "                                                                 \n",
            " activation_1 (Activation)   (None, 46)                0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 5,144,110\n",
            "Trainable params: 5,144,110\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Train Model"
      ],
      "metadata": {
        "id": "W5s8OD61h3NM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.callbacks import EarlyStopping\n",
        "es = EarlyStopping(monitor = 'val_loss',patience=3,verbose=1,mode='min')\n",
        "model.compile(optimizer ='adam',\n",
        "             loss= 'categorical_crossentropy',\n",
        "             metrics = ['accuracy'])\n",
        "h = model.fit(x_train, y_train,\n",
        "             epochs=100,\n",
        "             batch_size=32,\n",
        "             validation_split=0.1,\n",
        "             callbacks=[es])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1JR4WxQviH_r",
        "outputId": "f93f9c99-dd85-4ee9-b8b9-a0d593a5e1b9"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "253/253 [==============================] - 10s 37ms/step - loss: 1.2827 - accuracy: 0.7223 - val_loss: 0.9460 - val_accuracy: 0.7976\n",
            "Epoch 2/100\n",
            "253/253 [==============================] - 9s 36ms/step - loss: 0.4923 - accuracy: 0.8857 - val_loss: 0.8625 - val_accuracy: 0.8187\n",
            "Epoch 3/100\n",
            "253/253 [==============================] - 9s 34ms/step - loss: 0.2834 - accuracy: 0.9360 - val_loss: 0.9413 - val_accuracy: 0.8053\n",
            "Epoch 4/100\n",
            "253/253 [==============================] - 12s 48ms/step - loss: 0.2225 - accuracy: 0.9475 - val_loss: 0.9338 - val_accuracy: 0.8098\n",
            "Epoch 5/100\n",
            "253/253 [==============================] - 13s 51ms/step - loss: 0.2005 - accuracy: 0.9532 - val_loss: 1.0003 - val_accuracy: 0.8076\n",
            "Epoch 5: early stopping\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Evaluate Model"
      ],
      "metadata": {
        "id": "Cc94p1MDiMiT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.evaluate(x_test,y_test,batch_size=32,verbose=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HOer5Z37iPeG",
        "outputId": "88164917-1e45-44a1-8fa7-19c294f87e7a"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "71/71 [==============================] - 1s 11ms/step - loss: 0.9828 - accuracy: 0.7974\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.9828191995620728, 0.7974176406860352]"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Training and Validation Accuracy"
      ],
      "metadata": {
        "id": "EczRBDBJiMfU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(h.history['accuracy'],label='Training Accuracy')\n",
        "plt.plot(h.history['val_accuracy'],label = 'Validation Accuracy')\n",
        "plt.title('Taining and Validation Accuracy')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "XrjklkSHiQ5v",
        "outputId": "88593ddf-503f-4b80-cdca-8e6145ae47f6"
      },
      "execution_count": 7,
      "outputs": [
        {
          "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": "markdown",
      "source": [
        "### Training and Validation Loss"
      ],
      "metadata": {
        "id": "eKalyIRuiMcr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(h.history['loss'],label='Training Loss')\n",
        "plt.plot(h.history['val_loss'],label = 'Validation loss')\n",
        "plt.title('Taining and Validation Loss')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "6PeSOztwiZXC",
        "outputId": "4547ae66-1cfd-4072-9482-8ab53a545f27"
      },
      "execution_count": 8,
      "outputs": [
        {
          "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": "markdown",
      "source": [
        "##"
      ],
      "metadata": {
        "id": "PfU21YtkiMac"
      }
    }
  ]
}