DeepLearning_Sentdex.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Sentdex - Deep learning with neural networks\n",
    "import os\n",
    "import tensorflow as tf\n",
    "\n",
    "x1 = tf.constant(5)\n",
    "x2 = tf.constant(6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30\n"
     ]
    }
   ],
   "source": [
    "#Multiply the constants\n",
    "result = tf.multiply(x1,x2)\n",
    "\n",
    "\n",
    "#Run the tensorflow session and check result\n",
    "sess = tf.Session()\n",
    "print(sess.run(result))\n",
    "\n",
    "#Close the session\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30\n"
     ]
    }
   ],
   "source": [
    "#It will automatically close the session when done - better way to do it\n",
    "with tf.Session() as sess:\n",
    "    print(sess.run(result))\n",
    "       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting /tmp/data\\train-images-idx3-ubyte.gz\n",
      "Extracting /tmp/data\\train-labels-idx1-ubyte.gz\n",
      "Extracting /tmp/data\\t10k-images-idx3-ubyte.gz\n",
      "Extracting /tmp/data\\t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "#MNIST - 60k training examples of handwritten digits (28*28 pixels)\n",
    "#10k testing examples\n",
    "\n",
    "#Objective : Take the examples and pass it to Neural network and output the correct the correct number\n",
    "#Features - Is there anything on the pixel or not?\n",
    "\n",
    "#input > weight > hidden layer 1 > (activation function) > weights > hidden layer 2\n",
    "#(activation function) > weights > output layer\n",
    "#Feed forward NN \n",
    "#Compare output to intended output > cost function\n",
    "#optimization function (optimizer) > minimize cost (AdamOptimizer..., AdaGrad)\n",
    "#backpropagation\n",
    "#feed forward + backprop = epoch\n",
    "\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "#one_hot parameter useful for multi-class classification\n",
    "mnist = input_data.read_data_sets(\"/tmp/data\",one_hot=True)\n",
    "#10 classes : 0-9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Three hidden layers\n",
    "n_nodes_hl1 = 1000\n",
    "n_nodes_hl2 = 1000\n",
    "n_nodes_hl3 = 1000\n",
    "\n",
    "#Number of classes\n",
    "n_classes = 10\n",
    "\n",
    "#Will go through 100 features at a time\n",
    "batch_size = 100\n",
    "\n",
    "#Placeholder variables (height * width)\n",
    "#These are placeholders for some values in the graph\n",
    "x = tf.placeholder('float',[None,784])\n",
    "y = tf.placeholder('float')\n",
    "\n",
    "def neural_network_model(data):\n",
    "    \n",
    "    hidden_1_layer = {'weights' : tf.Variable(tf.random_normal([784,n_nodes_hl1])),\n",
    "                     'biases' : tf.Variable(tf.random_normal([n_nodes_hl1]))}\n",
    "    #bias is used to make some neurons fire even if all inputs is 0\n",
    "    hidden_2_layer = {'weights' : tf.Variable(tf.random_normal([n_nodes_hl1,n_nodes_hl2])),\n",
    "                     'biases' : tf.Variable(tf.random_normal([n_nodes_hl2]))}\n",
    "    hidden_3_layer = {'weights' : tf.Variable(tf.random_normal([n_nodes_hl2,n_nodes_hl3])),\n",
    "                     'biases' : tf.Variable(tf.random_normal([n_nodes_hl3]))}\n",
    "    output_layer = {'weights' : tf.Variable(tf.random_normal([n_nodes_hl3,n_classes])),\n",
    "                     'biases' : tf.Variable(tf.random_normal([n_classes]))}\n",
    "    #(input_data*weights) + biases\n",
    "    l1 = tf.add(tf.matmul(data,hidden_1_layer['weights']),hidden_1_layer['biases'])\n",
    "    l1 = tf.nn.relu(l1)\n",
    "    \n",
    "    l2 = tf.add(tf.matmul(l1,hidden_2_layer['weights']),hidden_2_layer['biases'])\n",
    "    l2 = tf.nn.relu(l2)\n",
    "    \n",
    "    l3 = tf.add(tf.matmul(l2,hidden_3_layer['weights']),hidden_3_layer['biases'])\n",
    "    l3 = tf.nn.relu(l3)\n",
    "    \n",
    "    output = tf.matmul(l3,output_layer['weights'])+ output_layer['biases']\n",
    "    \n",
    "    return output\n",
    "\n",
    "#Boom - Modeled a neural network\n",
    "#Done with the computation graph\n",
    "#We have coded the model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Running our network\n",
    "def train_neural_network(x):\n",
    "    prediction = neural_network_model(x)\n",
    "    #Cost function is cross entropy with logits\n",
    "    cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=prediction,labels=y))\n",
    "    \n",
    "    #Choose the optimizer\n",
    "    optimizer = tf.train.AdamOptimizer().minimize(cost)\n",
    "    \n",
    "    #Cycles feed forward + backprop\n",
    "    hm_epochs = 10\n",
    "    \n",
    "    with tf.Session() as sess:\n",
    "        sess.run(tf.global_variables_initializer())\n",
    "        #Trainng the network\n",
    "        for epoch in range(hm_epochs):\n",
    "            epoch_loss = 0\n",
    "            for _ in range(int(mnist.train.num_examples/batch_size)):\n",
    "                epoch_x,epoch_y = mnist.train.next_batch(batch_size)\n",
    "                _, c = sess.run([optimizer,cost], feed_dict = {x:epoch_x,y:epoch_y})\n",
    "                epoch_loss += c\n",
    "            print('Epoch',epoch,'Completed out of',hm_epochs,'loss:',epoch_loss)\n",
    "                    \n",
    "        \n",
    "        correct = tf.equal(tf.argmax(prediction,1),tf.argmax(y,1))\n",
    "        accuracy = tf.reduce_mean(tf.cast(correct,'float'))\n",
    "        print('Accuracy:',accuracy.eval({x:mnist.test.images,y:mnist.test.labels}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From <ipython-input-7-963725452600>:5: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "\n",
      "Future major versions of TensorFlow will allow gradients to flow\n",
      "into the labels input on backprop by default.\n",
      "\n",
      "See tf.nn.softmax_cross_entropy_with_logits_v2.\n",
      "\n",
      "Epoch 0 Completed out of 10 loss: 3758147.80841\n",
      "Epoch 1 Completed out of 10 loss: 796936.213623\n",
      "Epoch 2 Completed out of 10 loss: 372440.355949\n",
      "Epoch 3 Completed out of 10 loss: 188040.790529\n",
      "Epoch 4 Completed out of 10 loss: 114344.558451\n",
      "Epoch 5 Completed out of 10 loss: 91873.0738357\n",
      "Epoch 6 Completed out of 10 loss: 85191.3208058\n",
      "Epoch 7 Completed out of 10 loss: 69750.7198949\n",
      "Epoch 8 Completed out of 10 loss: 67209.9048491\n",
      "Epoch 9 Completed out of 10 loss: 56794.7105298\n",
      "Accuracy: 0.9593\n"
     ]
    }
   ],
   "source": [
    "train_neural_network(x)"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}