Analysis using dense layers and Cnn.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5"
   },
   "outputs": [],
   "source": [
    "from keras import Model, Input\n",
    "from keras import models\n",
    "from keras import layers\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Load data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "_cell_guid": "79c7e3d0-c299-4dcb-8224-4455121ee9b0",
    "_uuid": "d629ff2d2480ee46fbb7e2d37f6b5fab8052498a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50000, 2)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"IMDB Dataset.csv\", names=['review', 'сlass'], header=0,encoding='utf-8')\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = data['review']\n",
    "y_train = data['сlass']\n",
    "del data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "One of the other reviewers has mentioned that after watching just 1 Oz episode you'll be hooked. They are right, as this is exactly what happened with me.<br /><br />The first thing that struck me about Oz was its brutality and unflinching scenes of violence, which set in right from the word GO. Trust me, this is not a show for the faint hearted or timid. This show pulls no punches with regards to drugs, sex or violence. Its is hardcore, in the classic use of the word.<br /><br />It is called OZ as that is the nickname given to the Oswald Maximum Security State Penitentary. It focuses mainly on Emerald City, an experimental section of the prison where all the cells have glass fronts and face inwards, so privacy is not high on the agenda. Em City is home to many..Aryans, Muslims, gangstas, Latinos, Christians, Italians, Irish and more....so scuffles, death stares, dodgy dealings and shady agreements are never far away.<br /><br />I would say the main appeal of the show is due to the fact that it goes where other shows wouldn't dare. Forget pretty pictures painted for mainstream audiences, forget charm, forget romance...OZ doesn't mess around. The first episode I ever saw struck me as so nasty it was surreal, I couldn't say I was ready for it, but as I watched more, I developed a taste for Oz, and got accustomed to the high levels of graphic violence. Not just violence, but injustice (crooked guards who'll be sold out for a nickel, inmates who'll kill on order and get away with it, well mannered, middle class inmates being turned into prison bitches due to their lack of street skills or prison experience) Watching Oz, you may become comfortable with what is uncomfortable viewing....thats if you can get in touch with your darker side.\n",
      "positive\n",
      "50000\n"
     ]
    }
   ],
   "source": [
    "print(x_train[0])\n",
    "print(y_train[0])\n",
    "print(len(x_train))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Prepare X data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_words = 20_000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokenizer = Tokenizer(num_words=max_words)\n",
    "tokenizer.fit_on_texts(x_train)\n",
    "x_train = tokenizer.texts_to_sequences(x_train)\n",
    "del tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def vectorize(sequences, dimension):\n",
    "    results = np.zeros((len(sequences), dimension))\n",
    "    for i, sequence in enumerate(sequences):\n",
    "        for j in sequence:\n",
    "            results[i, j] += 1\n",
    "    return results/np.amax(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = vectorize(x_train, max_words)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Prepare Y data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train = [1 if i == 'positive' else 0 for i in y_train]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Split data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test = x_train[:10_000]\n",
    "y_test = y_train[:10_000]\n",
    "\n",
    "x_train = x_train[10_000:]\n",
    "y_train = y_train[10_000:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test = np.array(x_test)\n",
    "y_test = np.array(y_test)\n",
    "\n",
    "x_train = np.array(x_train)\n",
    "y_train = np.array(y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Model 1**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Conv1D, MaxPooling1D, Dense, Flatten, Dropout\n",
    "from keras.optimizers import Adam\n",
    "from keras.callbacks import TensorBoard\n",
    "from keras.layers import LeakyReLU\n",
    "from keras.layers import BatchNormalization\n",
    "from tensorflow.keras import regularizers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Dense(100, activation='sigmoid', kernel_initializer='he_uniform'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(LeakyReLU(alpha=0.1))\n",
    "model.add(Dense(75, activation='sigmoid'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(LeakyReLU(alpha=0.1))\n",
    "model.add(Dense(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss ='binary_crossentropy', optimizer='rmsprop' ,metrics =['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Train model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "79/79 [==============================] - 384s 5s/step - loss: 0.6932 - acc: 0.5007\n",
      "Epoch 2/10\n",
      "79/79 [==============================] - 402s 5s/step - loss: 0.6932 - acc: 0.4981\n",
      "Epoch 3/10\n",
      "79/79 [==============================] - 409s 5s/step - loss: 0.6932 - acc: 0.5007\n",
      "Epoch 4/10\n",
      "79/79 [==============================] - 411s 5s/step - loss: 0.6932 - acc: 0.4995\n",
      "Epoch 5/10\n",
      "79/79 [==============================] - 415s 5s/step - loss: 0.6932 - acc: 0.4946\n",
      "Epoch 6/10\n",
      "79/79 [==============================] - 420s 5s/step - loss: 0.6932 - acc: 0.4994\n",
      "Epoch 7/10\n",
      "79/79 [==============================] - 416s 5s/step - loss: 0.6932 - acc: 0.4988\n",
      "Epoch 8/10\n",
      "79/79 [==============================] - 417s 5s/step - loss: 0.6932 - acc: 0.5001\n",
      "Epoch 9/10\n",
      "79/79 [==============================] - 418s 5s/step - loss: 0.6932 - acc: 0.4996\n",
      "Epoch 10/10\n",
      "79/79 [==============================] - 418s 5s/step - loss: 0.6932 - acc: 0.5007\n"
     ]
    }
   ],
   "source": [
    "epochs = 10\n",
    "\n",
    "history = model.fit(x_train, y_train, batch_size = 512, epochs = epochs, verbose = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Loss evolution')"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 8))\n",
    "\n",
    "plt.subplot(2, 2, 1)\n",
    "plt.plot(history.history['loss'], label='Loss')\n",
    "plt.legend()\n",
    "plt.title('Loss evolution')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Test model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 2ms/step - loss: 2.6877 - accuracy: 0.7466\n"
     ]
    }
   ],
   "source": [
    "scores = model.evaluate(x_test, y_test, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing.sequence import pad_sequences\n",
    "X_train = pad_sequences(x_train, maxlen = max_words)\n",
    "X_test = pad_sequences(x_test, maxlen = max_words)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Model 2**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# Building the Model\n",
    "max_features = 10000 \n",
    "model = models.Sequential()\n",
    "model.add(layers.Embedding(max_features, 32, input_length = max_words))\n",
    "model.add(layers.Conv1D(32, 5, activation = 'relu'))\n",
    "model.add(layers.MaxPooling1D(3))\n",
    "model.add(layers.Dropout(0.5))\n",
    "model.add(layers.Dense(1, activation = 'sigmoid'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss ='binary_crossentropy', optimizer='rmsprop' ,metrics =['accuracy'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_3\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_2 (Embedding)      (None, 20000, 32)         320000    \n",
      "_________________________________________________________________\n",
      "conv1d_4 (Conv1D)            (None, 19996, 32)         5152      \n",
      "_________________________________________________________________\n",
      "max_pooling1d_4 (MaxPooling1 (None, 6665, 32)          0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 6665, 32)          0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 6665, 1)           33        \n",
      "=================================================================\n",
      "Total params: 325,185\n",
      "Trainable params: 325,185\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Train model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "63/63 [==============================] - 324s 5s/step - loss: 0.6932 - acc: 0.5003 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 2/10\n",
      "63/63 [==============================] - 322s 5s/step - loss: 0.6932 - acc: 0.5010 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 3/10\n",
      "63/63 [==============================] - 319s 5s/step - loss: 0.6932 - acc: 0.5017 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 4/10\n",
      "63/63 [==============================] - 321s 5s/step - loss: 0.6931 - acc: 0.5017 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 5/10\n",
      "63/63 [==============================] - 321s 5s/step - loss: 0.6931 - acc: 0.5029 - val_loss: 0.6931 - val_acc: 0.5033\n",
      "Epoch 6/10\n",
      "63/63 [==============================] - 324s 5s/step - loss: 0.6932 - acc: 0.5017 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 7/10\n",
      "63/63 [==============================] - 326s 5s/step - loss: 0.6932 - acc: 0.5017 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 8/10\n",
      "63/63 [==============================] - 327s 5s/step - loss: 0.6931 - acc: 0.5017 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 9/10\n",
      "63/63 [==============================] - 326s 5s/step - loss: 0.6932 - acc: 0.5017 - val_loss: 0.6932 - val_acc: 0.4967\n",
      "Epoch 10/10\n",
      "63/63 [==============================] - 326s 5s/step - loss: 0.6932 - acc: 0.5017 - val_loss: 0.6932 - val_acc: 0.4967\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, batch_size = 512, epochs = 10, validation_split = 0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# Plotting Training and Validation Loss\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Loss evolution')"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 8))\n",
    "\n",
    "plt.subplot(2, 2, 1)\n",
    "plt.plot(history.history['loss'], label='Loss')\n",
    "plt.plot(history.history['val_loss'], label='val_Loss')\n",
    "plt.legend()\n",
    "plt.title('Loss evolution')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}