lstmNet.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Recurrent Neural Net with Embeddings\n",
    "Adapted from: https://github.com/fchollet/keras/blob/master/examples/imdb_lstm.py"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Trains an LSTM model on the IMDB sentiment classification task.\n",
    "The dataset is actually **too small for LSTM** to be of any advantage\n",
    "compared to simpler, much faster methods such as TF-IDF + Logistic Regression.\n",
    "### Notes\n",
    "- RNNs are tricky. Choice of batch size is important,\n",
    "choice of loss and optimizer is critical, etc.\n",
    "Some configurations won't converge.\n",
    "- LSTM loss decrease patterns during training can be quite different\n",
    "from what you see with CNNs/MLPs/etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.preprocessing import sequence\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Embedding, Dropout\n",
    "from tensorflow.keras.layers import LSTM\n",
    "from tensorflow.keras.datasets import imdb\n",
    "\n",
    "max_num_words = 10000\n",
    "maxlen = 80\n",
    "batch_size = 32\n",
    "embedding_dims = 50\n",
    "hidden_dims = 250\n",
    "epochs = 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sentiment Analysis\n",
    "With Sentiment Analysis, we want to determine the attitude (e.g the sentiment) of for example a speaker or writer with respect to a document, interaction, or event. Therefore it is a natural language processing problem where text needs to be understood, to predict the underlying intent. The sentiment is mostly categorized into positive, negative and neutral categories. With the use of Sentiment Analysis, we want to predict for example a customers opinion and attitude about a product based on a review he wrote about it. Because of that, Sentiment Analysis is widely applied to things like reviews, surveys, documents and much more."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The Imdb Dataset\n",
    "The imdb sentiment classification dataset consists of 50,000 movie reviews from imdb users that are labeled as either positive (1) or negative (0). The reviews are preprocessed and each one is encoded as a sequence of word indexes in the form of integers. The words within the reviews are indexed by their overall frequency within the dataset. For example, the integer “2” encodes the second most frequent word in the data. The 50,000 reviews are split into 25,000 for training and 25,000 for testing. The dataset was created by researchers of the Stanford University and published in a [paper in 2011](http://www.aclweb.org/anthology/P11-1015), where they achieved 88.89% accuracy. It was also used within the “Bag of Words Meets Bags of Popcorn” Kaggle competition in 2011."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading data...\n"
     ]
    }
   ],
   "source": [
    "print('Loading data...')\n",
    "(x_train, y_train), (x_test, y_test) = imdb.load_data(num_words=max_num_words)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25000 train sequences\n",
      "25000 test sequences\n",
      "Categories: [0 1]\n",
      "Number of unique words: 9998\n",
      "Average Review length: 238.71364\n",
      "Standard Deviation: 176\n"
     ]
    }
   ],
   "source": [
    "print(len(x_train), 'train sequences')\n",
    "print(len(x_test), 'test sequences')\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "print(\"Categories:\", np.unique(y_train))\n",
    "print(\"Number of unique words:\", len(np.unique(np.hstack(x_train))))\n",
    "length = [len(i) for i in x_train]\n",
    "print(\"Average Review length:\", np.mean(length))\n",
    "print(\"Standard Deviation:\", round(np.std(length)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The output shows that the dataset is labeled into two categories, either 0 or 1, which represent the sentiment of the review. The whole dataset contains 9998 unique words and the average review length is 234 words, with a standard deviation of 173 words.\n",
    "\n",
    "Now we will look at a single training example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Label: 1\n",
      "[1, 14, 22, 16, 43, 530, 973, 1622, 1385, 65, 458, 4468, 66, 3941, 4, 173, 36, 256, 5, 25, 100, 43, 838, 112, 50, 670, 2, 9, 35, 480, 284, 5, 150, 4, 172, 112, 167, 2, 336, 385, 39, 4, 172, 4536, 1111, 17, 546, 38, 13, 447, 4, 192, 50, 16, 6, 147, 2025, 19, 14, 22, 4, 1920, 4613, 469, 4, 22, 71, 87, 12, 16, 43, 530, 38, 76, 15, 13, 1247, 4, 22, 17, 515, 17, 12, 16, 626, 18, 2, 5, 62, 386, 12, 8, 316, 8, 106, 5, 4, 2223, 5244, 16, 480, 66, 3785, 33, 4, 130, 12, 16, 38, 619, 5, 25, 124, 51, 36, 135, 48, 25, 1415, 33, 6, 22, 12, 215, 28, 77, 52, 5, 14, 407, 16, 82, 2, 8, 4, 107, 117, 5952, 15, 256, 4, 2, 7, 3766, 5, 723, 36, 71, 43, 530, 476, 26, 400, 317, 46, 7, 4, 2, 1029, 13, 104, 88, 4, 381, 15, 297, 98, 32, 2071, 56, 26, 141, 6, 194, 7486, 18, 4, 226, 22, 21, 134, 476, 26, 480, 5, 144, 30, 5535, 18, 51, 36, 28, 224, 92, 25, 104, 4, 226, 65, 16, 38, 1334, 88, 12, 16, 283, 5, 16, 4472, 113, 103, 32, 15, 16, 5345, 19, 178, 32]\n"
     ]
    }
   ],
   "source": [
    "print(\"Label:\", y_train[0])\n",
    "print(x_train[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first review of the dataset is labeled as positive (1).\n",
    "The function `get_word_index()` retrieves the dictionary mapping words to indices. We invert the index so that we can get back the original words and we can read them. Unknown words are replaced with a “#”.\n",
    "\n",
    "`get_words_index()` starts numbering from 3."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "index = imdb.get_word_index()\n",
    "\n",
    "word_index = { w:(k+3) for w,k in index.items() if k < max_num_words}\n",
    "word_index['<PAD>'] = 0\n",
    "word_index['<START>'] = 1\n",
    "word_index['<UNK>'] = 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Preparation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pad the sequences so that they have all the same length."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train shape: (25000, 80)\n",
      "x_test shape: (25000, 80)\n"
     ]
    }
   ],
   "source": [
    "x_train = sequence.pad_sequences(x_train, maxlen=maxlen)\n",
    "x_test = sequence.pad_sequences(x_test, maxlen=maxlen)\n",
    "print('x_train shape:', x_train.shape)\n",
    "print('x_test shape:', x_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build the Model\n",
    "We can now build our simple Neural Network. We start by defining the type of model we want to build. There are two types of models available in Keras: the Sequential model and the Model class used with functional API.\n",
    "\n",
    "Then we simply add the input-, hidden- and output-layers.\n",
    "The input layer is an embedding layer, used to represent each word as a vector in a space of dimensions `hidden_dims`.\n",
    "\n",
    "The first hidden layer is an `LSTM` layer, to which we add a dropout to prevent overfitting.\n",
    "Note that you should always use a dropout rate between 20% and 50%.\n",
    "\n",
    "At the output layer, we use a Dense layer with the sigmoid as activation function, which maps the values between 0 and 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Embedding(max_num_words, embedding_dims))\n",
    "model.add(LSTM(hidden_dims, dropout=0.2, recurrent_dropout=0.2))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "# try using different optimizers and different optimizer configs\n",
    "model.compile(loss='binary_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Show the model we have just built."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from tensorflow.keras.utils import plot_model\n",
    "plot_model(model, show_shapes=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the Model\n",
    "We are now able to train our model. We do this with a batch_size of 32 and only for two epochs because I recognized that the model overfits if we train it longer.\n",
    "\n",
    "The Batch size defines the number of samples that will be propagated through the network and an epoch is an iteration over the entire training data. In general a larger batch-size results in faster training, but don’t always converges fast. A smaller batch-size is slower in training but it can converge faster.\n",
    "\n",
    "This is definitely problem dependent and you need to try out a few different values. If you start with a problem for the first time, I would you recommend to you to first use a batch-size of 32, which is the standard size."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2\n",
      "704/704 [==============================] - 189s 265ms/step - loss: 0.4571 - accuracy: 0.7788 - val_loss: 0.3738 - val_accuracy: 0.8352\n",
      "Epoch 2/2\n",
      "704/704 [==============================] - 187s 265ms/step - loss: 0.3000 - accuracy: 0.8749 - val_loss: 0.3979 - val_accuracy: 0.8288\n",
      "Train accuracy 0.8320000171661377\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x_train, y_train,\n",
    "          batch_size=batch_size,\n",
    "          epochs=epochs,\n",
    "          validation_split=0.1)\n",
    "print('Train accuracy', np.mean(history.history[\"val_accuracy\"]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25000/25000 [==============================] - 139s 6ms/step\n",
      "Test score: 0.44111163621902466\n",
      "Test accuracy: 0.8271600008010864\n"
     ]
    }
   ],
   "source": [
    "score, acc = model.evaluate(x_test, y_test,\n",
    "                            batch_size=batch_size)\n",
    "print('Test score:', score)\n",
    "print('Test accuracy:', acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bidirectional LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from keras.layers import Bidirectional\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Embedding(max_num_words, embedding_dims, input_length=maxlen))\n",
    "model.add(Bidirectional(LSTM(64)))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "plot_model(model, show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# try using different optimizers and different optimizer configs\n",
    "model.compile(loss='binary_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/indexed_slices.py:433: UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory.\n",
      "  \"Converting sparse IndexedSlices to a dense Tensor of unknown shape. \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 22500 samples, validate on 2500 samples\n",
      "Epoch 1/2\n",
      "22500/22500 [==============================] - 148s 7ms/step - loss: 0.4402 - accuracy: 0.7861 - val_loss: 0.3617 - val_accuracy: 0.8388\n",
      "Epoch 2/2\n",
      "22500/22500 [==============================] - 148s 7ms/step - loss: 0.2751 - accuracy: 0.8885 - val_loss: 0.3664 - val_accuracy: 0.8392\n",
      "Train accuracy 0.839000016450882\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x_train, y_train,\n",
    "          batch_size=batch_size,\n",
    "          epochs=epochs,\n",
    "          validation_split=0.1)\n",
    "print('Train accuracy', np.mean(history.history[\"val_accuracy\"]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25000/25000 [==============================] - 69s 3ms/step\n",
      "Test score: 0.37134366543769837\n",
      "Test accuracy: 0.8381999731063843\n"
     ]
    }
   ],
   "source": [
    "score, acc = model.evaluate(x_test, y_test,\n",
    "                            batch_size=batch_size)\n",
    "print('Test score:', score)\n",
    "print('Test accuracy:', acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# One-hot Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def vectorize(sequences, dimension = 10000):\n",
    "    results = np.zeros((len(sequences), dimension))\n",
    "    for i, sequence in enumerate(sequences):\n",
    "        results[i, sequence] += 1\n",
    "    return results\n",
    " \n",
    "vdata = vectorize(data)\n",
    "targets = np.array(targets).astype(\"float32\")\n",
    "\n",
    "test_x = vdata[:10000]\n",
    "test_y = targets[:10000]\n",
    "train_x = vdata[10000:]\n",
    "train_y = targets[10000:]\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Dense(50, activation = \"relu\", input_shape=(10000, )))\n",
    "# Hidden - Layers\n",
    "model.add(Dropout(0.3, noise_shape=None, seed=None))\n",
    "model.add(Dense(50, activation = \"relu\"))\n",
    "model.add(Dropout(0.2, noise_shape=None, seed=None))\n",
    "model.add(Dense(50, activation = \"relu\"))\n",
    "# Output- Layer\n",
    "model.add(Dense(1, activation = \"sigmoid\"))\n",
    "\n",
    "plot_model(model, show_shapes=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vdata[0][14]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 40000 samples, validate on 10000 samples\n",
      "Epoch 1/2\n",
      "40000/40000 [==============================] - 8s 207us/step - loss: 0.4180 - accuracy: 0.8119 - val_loss: 0.2643 - val_accuracy: 0.8930\n",
      "Epoch 2/2\n",
      "40000/40000 [==============================] - 8s 204us/step - loss: 0.2181 - accuracy: 0.9176 - val_loss: 0.2657 - val_accuracy: 0.8945\n",
      "Train Accuracy: 0.893750011920929\n"
     ]
    }
   ],
   "source": [
    "model.compile(loss='binary_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "history = model.fit(train_x, train_y,\n",
    "     epochs= 2,\n",
    "     batch_size = 500,\n",
    "     validation_split = 0.1\n",
    ")\n",
    "print(\"Train Accuracy:\", np.mean(history.history[\"val_accuracy\"]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With this simple model, we already beat the accuracy of the 2011 paper mentioned st the beginning.\n",
    "Feel free to experiment with the hyperparameters and the number of layers."
   ]
  }
 ],
 "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}