Basic_MLP_combined_comparison.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000 train samples\n",
      "10000 test samples\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_1 (Dense)              (None, 512)               401920    \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 10)                5130      \n",
      "=================================================================\n",
      "Total params: 669,706\n",
      "Trainable params: 669,706\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/20\n",
      "60000/60000 [==============================] - 3s - loss: 0.2481 - acc: 0.9246 - val_loss: 0.1113 - val_acc: 0.9660\n",
      "Epoch 2/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.1034 - acc: 0.9691 - val_loss: 0.0788 - val_acc: 0.9754\n",
      "Epoch 3/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0747 - acc: 0.9773 - val_loss: 0.0849 - val_acc: 0.9749\n",
      "Epoch 4/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0604 - acc: 0.9817 - val_loss: 0.0652 - val_acc: 0.9819\n",
      "Epoch 5/20\n",
      "60000/60000 [==============================] - 1s - loss: 0.0494 - acc: 0.9851 - val_loss: 0.0861 - val_acc: 0.9792\n",
      "Epoch 6/20\n",
      "60000/60000 [==============================] - 1s - loss: 0.0426 - acc: 0.9871 - val_loss: 0.0984 - val_acc: 0.9773\n",
      "Epoch 7/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0381 - acc: 0.9882 - val_loss: 0.0813 - val_acc: 0.9817\n",
      "Epoch 8/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0351 - acc: 0.9899 - val_loss: 0.0899 - val_acc: 0.9811\n",
      "Epoch 9/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0306 - acc: 0.9913 - val_loss: 0.0906 - val_acc: 0.9822\n",
      "Epoch 10/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0293 - acc: 0.9918 - val_loss: 0.0885 - val_acc: 0.9825\n",
      "Epoch 11/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0262 - acc: 0.9922 - val_loss: 0.0841 - val_acc: 0.9834\n",
      "Epoch 12/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0248 - acc: 0.9930 - val_loss: 0.1085 - val_acc: 0.9809\n",
      "Epoch 13/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0224 - acc: 0.9936 - val_loss: 0.1024 - val_acc: 0.9813\n",
      "Epoch 14/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0225 - acc: 0.9940 - val_loss: 0.0988 - val_acc: 0.9834\n",
      "Epoch 15/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0212 - acc: 0.9941 - val_loss: 0.0976 - val_acc: 0.9832\n",
      "Epoch 16/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0229 - acc: 0.9940 - val_loss: 0.0888 - val_acc: 0.9842\n",
      "Epoch 17/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0190 - acc: 0.9951 - val_loss: 0.1014 - val_acc: 0.9836\n",
      "Epoch 18/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0199 - acc: 0.9949 - val_loss: 0.1069 - val_acc: 0.9842\n",
      "Epoch 19/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0185 - acc: 0.9949 - val_loss: 0.1303 - val_acc: 0.9812\n",
      "Epoch 20/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0191 - acc: 0.9951 - val_loss: 0.1033 - val_acc: 0.9846\n",
      "Test loss: 0.103326279387\n",
      "Test accuracy: 0.9846\n"
     ]
    }
   ],
   "source": [
    "'''Trains a simple deep NN on the MNIST dataset.\n",
    "\n",
    "Gets to 98.40% test accuracy after 20 epochs\n",
    "(there is *a lot* of margin for parameter tuning).\n",
    "2 seconds per epoch on a K520 GPU.\n",
    "\n",
    "\n",
    "'''\n",
    "\n",
    "from __future__ import print_function\n",
    "\n",
    "import keras\n",
    "from keras.datasets import mnist\n",
    "from keras.models import Sequential\n",
    "from keras.layers.core import Dense, Dropout, Activation\n",
    "from keras.layers.noise import AlphaDropout\n",
    "from keras.utils import np_utils\n",
    "from keras.optimizers import RMSprop\n",
    "\n",
    "batch_size = 128\n",
    "num_classes = 10\n",
    "epochs = 20\n",
    "\n",
    "# the data, shuffled and split between train and test sets\n",
    "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
    "\n",
    "x_train = x_train.reshape(60000, 784)\n",
    "x_test = x_test.reshape(10000, 784)\n",
    "x_train = x_train.astype('float32')\n",
    "x_test = x_test.astype('float32')\n",
    "x_train /= 255\n",
    "x_test /= 255\n",
    "print(x_train.shape[0], 'train samples')\n",
    "print(x_test.shape[0], 'test samples')\n",
    "\n",
    "# convert class vectors to binary class matrices\n",
    "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
    "y_test = keras.utils.to_categorical(y_test, num_classes)\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Dense(512, activation='relu', input_shape=(784,)))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(512, activation='relu'))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(10, activation='softmax'))\n",
    "\n",
    "model.summary()\n",
    "\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer=RMSprop(),\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "historyMLP = model.fit(x_train, y_train,\n",
    "                    batch_size=batch_size,\n",
    "                    epochs=epochs,\n",
    "                    verbose=1,\n",
    "                    validation_data=(x_test, y_test))\n",
    "scoreMLP = model.evaluate(x_test, y_test, verbose=0)\n",
    "print('Test loss:', scoreMLP[0])\n",
    "print('Test accuracy:', scoreMLP[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000 train samples\n",
      "10000 test samples\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_4 (Dense)              (None, 512)               401920    \n",
      "_________________________________________________________________\n",
      "alpha_dropout_1 (AlphaDropou (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "alpha_dropout_2 (AlphaDropou (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 10)                5130      \n",
      "=================================================================\n",
      "Total params: 669,706\n",
      "Trainable params: 669,706\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.4296 - acc: 0.8694 - val_loss: 0.2557 - val_acc: 0.9315\n",
      "Epoch 2/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.1921 - acc: 0.9419 - val_loss: 0.1701 - val_acc: 0.9568\n",
      "Epoch 3/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.1301 - acc: 0.9599 - val_loss: 0.1126 - val_acc: 0.9704\n",
      "Epoch 4/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.1013 - acc: 0.9687 - val_loss: 0.1257 - val_acc: 0.9673\n",
      "Epoch 5/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0865 - acc: 0.9736 - val_loss: 0.0985 - val_acc: 0.9747\n",
      "Epoch 6/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0720 - acc: 0.9781 - val_loss: 0.0926 - val_acc: 0.9769\n",
      "Epoch 7/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0633 - acc: 0.9803 - val_loss: 0.1023 - val_acc: 0.9763\n",
      "Epoch 8/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0559 - acc: 0.9825 - val_loss: 0.0976 - val_acc: 0.9790\n",
      "Epoch 9/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0508 - acc: 0.9844 - val_loss: 0.1148 - val_acc: 0.9761\n",
      "Epoch 10/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0460 - acc: 0.9863 - val_loss: 0.1059 - val_acc: 0.9796\n",
      "Epoch 11/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0422 - acc: 0.9871 - val_loss: 0.0885 - val_acc: 0.9802\n",
      "Epoch 12/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0386 - acc: 0.9880 - val_loss: 0.1305 - val_acc: 0.9759\n",
      "Epoch 13/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0350 - acc: 0.9889 - val_loss: 0.1177 - val_acc: 0.9783\n",
      "Epoch 14/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0305 - acc: 0.9900 - val_loss: 0.1174 - val_acc: 0.9809\n",
      "Epoch 15/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0308 - acc: 0.9909 - val_loss: 0.1155 - val_acc: 0.9809\n",
      "Epoch 16/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0281 - acc: 0.9914 - val_loss: 0.1183 - val_acc: 0.9828\n",
      "Epoch 17/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0264 - acc: 0.9921 - val_loss: 0.1133 - val_acc: 0.9823\n",
      "Epoch 18/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0217 - acc: 0.9934 - val_loss: 0.1152 - val_acc: 0.9826\n",
      "Epoch 19/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0231 - acc: 0.9929 - val_loss: 0.1198 - val_acc: 0.9820\n",
      "Epoch 20/20\n",
      "60000/60000 [==============================] - 2s - loss: 0.0236 - acc: 0.9931 - val_loss: 0.1281 - val_acc: 0.9818\n",
      "Test loss: 0.128061375022\n",
      "Test accuracy: 0.9818\n"
     ]
    }
   ],
   "source": [
    "'''Trains a simple SELU based deep NN on the MNIST dataset.\n",
    "\n",
    "Gets to 98.40% test accuracy after 20 epochs\n",
    "(there is *a lot* of margin for parameter tuning).\n",
    "2 seconds per epoch on a K520 GPU.\n",
    "\n",
    "\n",
    "'''\n",
    "\n",
    "from __future__ import print_function\n",
    "\n",
    "import keras\n",
    "from keras.datasets import mnist\n",
    "from keras.models import Sequential\n",
    "from keras.layers.core import Dense, Dropout, Activation\n",
    "from keras.layers.noise import AlphaDropout\n",
    "from keras.utils import np_utils\n",
    "from keras.optimizers import RMSprop, Adam\n",
    "\n",
    "batch_size = 128\n",
    "num_classes = 10\n",
    "epochs = 20\n",
    "learning_rate = 0.001\n",
    "\n",
    "# the data, shuffled and split between train and test sets\n",
    "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
    "\n",
    "x_train = x_train.reshape(60000, 784)\n",
    "x_test = x_test.reshape(10000, 784)\n",
    "x_train = x_train.astype('float32')\n",
    "x_test = x_test.astype('float32')\n",
    "x_train /= 255\n",
    "x_test /= 255\n",
    "print(x_train.shape[0], 'train samples')\n",
    "print(x_test.shape[0], 'test samples')\n",
    "\n",
    "# convert class vectors to binary class matrices\n",
    "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
    "y_test = keras.utils.to_categorical(y_test, num_classes)\n",
    "\n",
    "modelSELU = Sequential()\n",
    "modelSELU.add(Dense(512, activation='selu', input_shape=(784,)))\n",
    "modelSELU.add(AlphaDropout(0.1))\n",
    "modelSELU.add(Dense(512, activation='selu'))\n",
    "modelSELU.add(AlphaDropout(0.1))\n",
    "modelSELU.add(Dense(10, activation='softmax'))\n",
    "\n",
    "modelSELU.summary()\n",
    "\n",
    "modelSELU.compile(loss='categorical_crossentropy',\n",
    "              optimizer=RMSprop(),\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "historySELU = modelSELU.fit(x_train, y_train,\n",
    "                    batch_size=batch_size,\n",
    "                    epochs=epochs,\n",
    "                    verbose=1,\n",
    "                    validation_data=(x_test, y_test))\n",
    "scoreSELU = modelSELU.evaluate(x_test, y_test, verbose=0)\n",
    "print('Test loss:', scoreSELU[0])\n",
    "print('Test accuracy:', scoreSELU[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJkAAAJcCAYAAABaP3UWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f71e163f650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pylab\n",
    "\n",
    "plt.figure(figsize=(20,10))\n",
    "plt.plot(historyMLP.history['acc'])\n",
    "plt.plot(historyMLP.history['val_acc'])\n",
    "plt.plot(historySELU.history['acc'])\n",
    "plt.plot(historySELU.history['val_acc'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['MLP-train', 'MLP-test','SELU-train', 'SELU-test'], loc='lower right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJIAAAJcCAYAAACi347hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nERERERER1YRBEhERERERERER1eT/AqlP42UqTpAfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f72d447f510>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.plot(historyMLP.history['loss'])\n",
    "plt.plot(historyMLP.history['val_loss'])\n",
    "plt.plot(historySELU.history['loss'])\n",
    "plt.plot(historySELU.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['MLP-train', 'MLP-test', 'SELU-train', 'SELU-test'], loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting MNIST_data/train-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data/train-labels-idx1-ubyte.gz\n",
      "Extracting MNIST_data/t10k-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data/t10k-labels-idx1-ubyte.gz\n",
      "Epoch: 0001 cost = 0.520059154\n",
      "Epoch: 0002 cost = 0.161047111\n",
      "Epoch: 0003 cost = 0.132361550\n",
      "Epoch: 0004 cost = 0.106143619\n",
      "Epoch: 0005 cost = 0.096453789\n",
      "Epoch: 0006 cost = 0.096108608\n",
      "Epoch: 0007 cost = 0.081616417\n",
      "Epoch: 0008 cost = 0.076672313\n",
      "Epoch: 0009 cost = 0.080843923\n",
      "Epoch: 0010 cost = 0.073859843\n",
      "Epoch: 0011 cost = 0.061821199\n",
      "Epoch: 0012 cost = 0.064553186\n",
      "Epoch: 0013 cost = 0.061651825\n",
      "Epoch: 0014 cost = 0.049234065\n",
      "Epoch: 0015 cost = 0.051621947\n",
      "Epoch: 0016 cost = 0.053794841\n",
      "Epoch: 0017 cost = 0.049004902\n",
      "Epoch: 0018 cost = 0.045878440\n",
      "Epoch: 0019 cost = 0.043574250\n",
      "Epoch: 0020 cost = 0.054216822\n",
      "Epoch: 0021 cost = 0.036742025\n",
      "Epoch: 0022 cost = 0.039176778\n",
      "Epoch: 0023 cost = 0.040113735\n",
      "Epoch: 0024 cost = 0.040813351\n",
      "Epoch: 0025 cost = 0.032637658\n",
      "Epoch: 0026 cost = 0.081223909\n",
      "Epoch: 0027 cost = 0.030257178\n",
      "Epoch: 0028 cost = 0.020022301\n",
      "Epoch: 0029 cost = 0.035565121\n",
      "Epoch: 0030 cost = 0.024361674\n",
      "Epoch: 0031 cost = 0.049606615\n",
      "Epoch: 0032 cost = 0.026361921\n",
      "Epoch: 0033 cost = 0.027691311\n",
      "Epoch: 0034 cost = 0.032607261\n",
      "Epoch: 0035 cost = 0.034171752\n",
      "Epoch: 0036 cost = 0.026146087\n",
      "Epoch: 0037 cost = 0.018694299\n",
      "Epoch: 0038 cost = 0.046192122\n",
      "Epoch: 0039 cost = 0.035283037\n",
      "Epoch: 0040 cost = 0.024799137\n",
      "Epoch: 0041 cost = 0.031563395\n",
      "Epoch: 0042 cost = 0.028838067\n",
      "Epoch: 0043 cost = 0.022196829\n",
      "Epoch: 0044 cost = 0.028412000\n",
      "Epoch: 0045 cost = 0.038151027\n",
      "Epoch: 0046 cost = 0.019609048\n",
      "Epoch: 0047 cost = 0.019822850\n",
      "Epoch: 0048 cost = 0.034508349\n",
      "Epoch: 0049 cost = 0.019575027\n",
      "Epoch: 0050 cost = 0.033447701\n",
      "Learning Finished!\n",
      "Accuracy: 0.9764\n",
      "Label:  [2]\n",
      "Prediction:  [2]\n"
     ]
    }
   ],
   "source": [
    "# SELU implementation from https://github.com/bioinf-jku/SNNs/blob/master/selu.py\n",
    "import tensorflow as tf\n",
    "import random\n",
    "# import matplotlib.pyplot as plt\n",
    "# -*- coding: utf-8 -*-\n",
    "'''\n",
    "Tensorflow Implementation of the Scaled ELU function and Dropout\n",
    "'''\n",
    "\n",
    "import numbers\n",
    "from tensorflow.contrib import layers\n",
    "from tensorflow.python.framework import ops\n",
    "from tensorflow.python.framework import tensor_shape\n",
    "from tensorflow.python.framework import tensor_util\n",
    "from tensorflow.python.ops import math_ops\n",
    "from tensorflow.python.ops import random_ops\n",
    "from tensorflow.python.ops import array_ops\n",
    "from tensorflow.python.layers import utils\n",
    "\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "\n",
    "tf.set_random_seed(777)  # reproducibility\n",
    "\n",
    "def selu(x):\n",
    "    with ops.name_scope('elu') as scope:\n",
    "        alpha = 1.6732632423543772848170429916717\n",
    "        scale = 1.0507009873554804934193349852946\n",
    "        return scale*tf.where(x>=0.0, x, alpha*tf.nn.elu(x))\n",
    "\n",
    "\n",
    "def dropout_selu(x, keep_prob, alpha= -1.7580993408473766, fixedPointMean=0.0, fixedPointVar=1.0,\n",
    "                 noise_shape=None, seed=None, name=None, training=False):\n",
    "    \"\"\"Dropout to a value with rescaling.\"\"\"\n",
    "\n",
    "    def dropout_selu_impl(x, rate, alpha, noise_shape, seed, name):\n",
    "        keep_prob = 1.0 - rate\n",
    "        x = ops.convert_to_tensor(x, name=\"x\")\n",
    "        if isinstance(keep_prob, numbers.Real) and not 0 < keep_prob <= 1:\n",
    "            raise ValueError(\"keep_prob must be a scalar tensor or a float in the \"\n",
    "                                             \"range (0, 1], got %g\" % keep_prob)\n",
    "        keep_prob = ops.convert_to_tensor(keep_prob, dtype=x.dtype, name=\"keep_prob\")\n",
    "        keep_prob.get_shape().assert_is_compatible_with(tensor_shape.scalar())\n",
    "\n",
    "        alpha = ops.convert_to_tensor(alpha, dtype=x.dtype, name=\"alpha\")\n",
    "        keep_prob.get_shape().assert_is_compatible_with(tensor_shape.scalar())\n",
    "\n",
    "        if tensor_util.constant_value(keep_prob) == 1:\n",
    "            return x\n",
    "\n",
    "        noise_shape = noise_shape if noise_shape is not None else array_ops.shape(x)\n",
    "        random_tensor = keep_prob\n",
    "        random_tensor += random_ops.random_uniform(noise_shape, seed=seed, dtype=x.dtype)\n",
    "        binary_tensor = math_ops.floor(random_tensor)\n",
    "        ret = x * binary_tensor + alpha * (1-binary_tensor)\n",
    "\n",
    "        a = tf.sqrt(fixedPointVar / (keep_prob *((1-keep_prob) * tf.pow(alpha-fixedPointMean,2) + fixedPointVar)))\n",
    "\n",
    "        b = fixedPointMean - a * (keep_prob * fixedPointMean + (1 - keep_prob) * alpha)\n",
    "        ret = a * ret + b\n",
    "        ret.set_shape(x.get_shape())\n",
    "        return ret\n",
    "\n",
    "    with ops.name_scope(name, \"dropout\", [x]) as name:\n",
    "        return utils.smart_cond(training,\n",
    "                                lambda: dropout_selu_impl(x, keep_prob, alpha, noise_shape, seed, name),\n",
    "                                lambda: array_ops.identity(x))\n",
    "\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "# Check out https://www.tensorflow.org/get_started/mnist/beginners for\n",
    "# more information about the mnist dataset\n",
    "\n",
    "# parameters\n",
    "learning_rate = 0.001\n",
    "training_epochs = 50\n",
    "batch_size = 100\n",
    "\n",
    "# input place holders\n",
    "X = tf.placeholder(tf.float32, [None, 784])\n",
    "Y = tf.placeholder(tf.float32, [None, 10])\n",
    "\n",
    "# dropout (keep_prob) rate  0.7 on training, but should be 1 for testing\n",
    "keep_prob = tf.placeholder(tf.float32)\n",
    "\n",
    "# weights & bias for nn layers\n",
    "# http://stackoverflow.com/questions/33640581/how-to-do-xavier-initialization-on-tensorflow\n",
    "W1 = tf.get_variable(\"W1\", shape=[784, 512],\n",
    "                     initializer=tf.contrib.layers.xavier_initializer())\n",
    "b1 = tf.Variable(tf.random_normal([512]))\n",
    "L1 = selu(tf.matmul(X, W1) + b1)\n",
    "L1 = dropout_selu(L1, keep_prob=keep_prob)\n",
    "\n",
    "W2 = tf.get_variable(\"W2\", shape=[512, 512],\n",
    "                     initializer=tf.contrib.layers.xavier_initializer())\n",
    "b2 = tf.Variable(tf.random_normal([512]))\n",
    "L2 = selu(tf.matmul(L1, W2) + b2)\n",
    "L2 = dropout_selu(L2, keep_prob=keep_prob)\n",
    "\n",
    "W3 = tf.get_variable(\"W3\", shape=[512, 512],\n",
    "                     initializer=tf.contrib.layers.xavier_initializer())\n",
    "b3 = tf.Variable(tf.random_normal([512]))\n",
    "L3 = selu(tf.matmul(L2, W3) + b3)\n",
    "L3 = dropout_selu(L3, keep_prob=keep_prob)\n",
    "\n",
    "W4 = tf.get_variable(\"W4\", shape=[512, 512],\n",
    "                     initializer=tf.contrib.layers.xavier_initializer())\n",
    "b4 = tf.Variable(tf.random_normal([512]))\n",
    "L4 = selu(tf.matmul(L3, W4) + b4)\n",
    "L4 = dropout_selu(L4, keep_prob=keep_prob)\n",
    "\n",
    "W5 = tf.get_variable(\"W5\", shape=[512, 10],\n",
    "                     initializer=tf.contrib.layers.xavier_initializer())\n",
    "b5 = tf.Variable(tf.random_normal([10]))\n",
    "hypothesis = tf.matmul(L4, W5) + b5\n",
    "\n",
    "# define cost/loss & optimizer\n",
    "cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(\n",
    "    logits=hypothesis, labels=Y))\n",
    "optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(cost)\n",
    "\n",
    "# initialize\n",
    "sess = tf.Session()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "\n",
    "# train my model\n",
    "for epoch in range(training_epochs):\n",
    "    avg_cost = 0\n",
    "    total_batch = int(mnist.train.num_examples / batch_size)\n",
    "\n",
    "    for i in range(total_batch):\n",
    "        batch_xs, batch_ys = mnist.train.next_batch(batch_size)\n",
    "        feed_dict = {X: batch_xs, Y: batch_ys, keep_prob: 0.7}\n",
    "        c, _ = sess.run([cost, optimizer], feed_dict=feed_dict)\n",
    "        avg_cost += c / total_batch\n",
    "\n",
    "    print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.9f}'.format(avg_cost))\n",
    "\n",
    "print('Learning Finished!')\n",
    "\n",
    "# Test model and check accuracy\n",
    "correct_prediction = tf.equal(tf.argmax(hypothesis, 1), tf.argmax(Y, 1))\n",
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n",
    "print('Accuracy:', sess.run(accuracy, feed_dict={\n",
    "      X: mnist.test.images, Y: mnist.test.labels, keep_prob: 1}))\n",
    "\n",
    "# Get one and predict\n",
    "r = random.randint(0, mnist.test.num_examples - 1)\n",
    "print(\"Label: \", sess.run(tf.argmax(mnist.test.labels[r:r + 1], 1)))\n",
    "print(\"Prediction: \", sess.run(\n",
    "    tf.argmax(hypothesis, 1), feed_dict={X: mnist.test.images[r:r + 1], keep_prob: 1}))\n",
    "\n",
    "# plt.imshow(mnist.test.images[r:r + 1].\n",
    "#           reshape(28, 28), cmap='Greys', interpolation='nearest')\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading data...\n",
      "Downloading data from https://s3.amazonaws.com/text-datasets/reuters.npz\n",
      "1949696/2110848 [==========================>...] - ETA: 0s8982 train sequences\n",
      "2246 test sequences\n",
      "46 classes\n",
      "Vectorizing sequence data...\n",
      "x_train shape: (8982, 1000)\n",
      "x_test shape: (2246, 1000)\n",
      "Convert class vector to binary class matrix (for use with categorical_crossentropy)\n",
      "y_train shape: (8982, 46)\n",
      "y_test shape: (2246, 46)\n",
      "Building relu model...\n",
      "Train on 8083 samples, validate on 899 samples\n",
      "Epoch 1/25\n",
      "8083/8083 [==============================] - 1s - loss: 1.4220 - acc: 0.6801 - val_loss: 1.0807 - val_acc: 0.7608\n",
      "Epoch 2/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.7710 - acc: 0.8209 - val_loss: 0.9300 - val_acc: 0.7842\n",
      "Epoch 3/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.5456 - acc: 0.8661 - val_loss: 0.8650 - val_acc: 0.7987\n",
      "Epoch 4/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.4122 - acc: 0.8966 - val_loss: 0.8691 - val_acc: 0.8065\n",
      "Epoch 5/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.3266 - acc: 0.9155 - val_loss: 0.8979 - val_acc: 0.8053\n",
      "Epoch 6/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.2670 - acc: 0.9308 - val_loss: 0.9331 - val_acc: 0.8120\n",
      "Epoch 7/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.2290 - acc: 0.9391 - val_loss: 0.9448 - val_acc: 0.8042\n",
      "Epoch 8/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.2170 - acc: 0.9416 - val_loss: 0.9733 - val_acc: 0.8042\n",
      "Epoch 9/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1998 - acc: 0.9456 - val_loss: 1.0235 - val_acc: 0.8065\n",
      "Epoch 10/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1895 - acc: 0.9474 - val_loss: 1.0069 - val_acc: 0.7976\n",
      "Epoch 11/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1821 - acc: 0.9499 - val_loss: 1.0175 - val_acc: 0.7976\n",
      "Epoch 12/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1803 - acc: 0.9501 - val_loss: 1.0726 - val_acc: 0.8031\n",
      "Epoch 13/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1680 - acc: 0.9516 - val_loss: 1.0887 - val_acc: 0.7976\n",
      "Epoch 14/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1654 - acc: 0.9539 - val_loss: 1.0720 - val_acc: 0.7942\n",
      "Epoch 15/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1506 - acc: 0.9557 - val_loss: 1.1195 - val_acc: 0.7976\n",
      "Epoch 16/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1515 - acc: 0.9565 - val_loss: 1.1334 - val_acc: 0.7842\n",
      "Epoch 17/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1532 - acc: 0.9573 - val_loss: 1.1220 - val_acc: 0.8009\n",
      "Epoch 18/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1547 - acc: 0.9574 - val_loss: 1.1482 - val_acc: 0.7920\n",
      "Epoch 19/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1539 - acc: 0.9574 - val_loss: 1.1434 - val_acc: 0.7987\n",
      "Epoch 20/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.1504 - acc: 0.9588 - val_loss: 1.1585 - val_acc: 0.7887\n",
      "Epoch 21/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1324 - acc: 0.9600 - val_loss: 1.1936 - val_acc: 0.7931\n",
      "Epoch 22/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1467 - acc: 0.9583 - val_loss: 1.2086 - val_acc: 0.7909\n",
      "Epoch 23/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.1492 - acc: 0.9568 - val_loss: 1.1826 - val_acc: 0.7964\n",
      "Epoch 24/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.1437 - acc: 0.9595 - val_loss: 1.1799 - val_acc: 0.8020\n",
      "Epoch 25/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.1462 - acc: 0.9592 - val_loss: 1.2485 - val_acc: 0.7864\n",
      "2080/2246 [==========================>...] - ETA: 0sRELU model results\n",
      "Test score: 1.26111637368\n",
      "Test accuracy: 0.783615316171\n",
      "Building selu model...\n",
      "Train on 8083 samples, validate on 899 samples\n",
      "Epoch 1/25\n",
      "8083/8083 [==============================] - 1s - loss: 1.6095 - acc: 0.6350 - val_loss: 1.5445 - val_acc: 0.7597\n",
      "Epoch 2/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.9619 - acc: 0.7769 - val_loss: 1.3541 - val_acc: 0.7875\n",
      "Epoch 3/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.7587 - acc: 0.8207 - val_loss: 1.2169 - val_acc: 0.8009\n",
      "Epoch 4/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.6282 - acc: 0.8498 - val_loss: 1.3181 - val_acc: 0.7898\n",
      "Epoch 5/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.5344 - acc: 0.8673 - val_loss: 1.3203 - val_acc: 0.8120\n",
      "Epoch 6/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.4592 - acc: 0.8858 - val_loss: 1.3199 - val_acc: 0.8076\n",
      "Epoch 7/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.4117 - acc: 0.8947 - val_loss: 1.3126 - val_acc: 0.8109\n",
      "Epoch 8/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.3735 - acc: 0.9014 - val_loss: 1.4049 - val_acc: 0.8042\n",
      "Epoch 9/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.3361 - acc: 0.9128 - val_loss: 1.4440 - val_acc: 0.8176\n",
      "Epoch 10/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.3094 - acc: 0.9191 - val_loss: 1.5091 - val_acc: 0.8098\n",
      "Epoch 11/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.2831 - acc: 0.9234 - val_loss: 1.5243 - val_acc: 0.7976\n",
      "Epoch 12/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.2686 - acc: 0.9254 - val_loss: 1.5310 - val_acc: 0.8087\n",
      "Epoch 13/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.2486 - acc: 0.9306 - val_loss: 1.6591 - val_acc: 0.8031\n",
      "Epoch 14/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.2362 - acc: 0.9313 - val_loss: 1.6322 - val_acc: 0.7998\n",
      "Epoch 15/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.2243 - acc: 0.9325 - val_loss: 1.7275 - val_acc: 0.7942\n",
      "Epoch 16/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.2175 - acc: 0.9358 - val_loss: 1.6775 - val_acc: 0.8042\n",
      "Epoch 17/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1985 - acc: 0.9411 - val_loss: 1.7584 - val_acc: 0.8009\n",
      "Epoch 18/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.1914 - acc: 0.9442 - val_loss: 1.7898 - val_acc: 0.7909\n",
      "Epoch 19/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1871 - acc: 0.9441 - val_loss: 1.8381 - val_acc: 0.8020\n",
      "Epoch 20/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1806 - acc: 0.9451 - val_loss: 1.8494 - val_acc: 0.7920\n",
      "Epoch 21/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1826 - acc: 0.9447 - val_loss: 1.8112 - val_acc: 0.7931\n",
      "Epoch 22/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1731 - acc: 0.9448 - val_loss: 1.8369 - val_acc: 0.7931\n",
      "Epoch 23/25\n",
      "8083/8083 [==============================] - 0s - loss: 0.1670 - acc: 0.9483 - val_loss: 1.8642 - val_acc: 0.7931\n",
      "Epoch 24/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1678 - acc: 0.9494 - val_loss: 1.9149 - val_acc: 0.8009\n",
      "Epoch 25/25\n",
      "8083/8083 [==============================] - 1s - loss: 0.1593 - acc: 0.9524 - val_loss: 1.8989 - val_acc: 0.7987\n",
      "1920/2246 [========================>.....] - ETA: 0sSELU model results\n",
      "Test score: 2.0712619677\n",
      "Test accuracy: 0.780498664292\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEKCAYAAAAB0GKPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f725e057c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''\n",
    "\n",
    "Compare the performance of a simple MLP using two\n",
    "different activation functions: RELU and SELU\n",
    "on the Reuters newswire topic classification task.\n",
    "\n",
    "'''\n",
    "from __future__ import print_function\n",
    "\n",
    "import numpy as np\n",
    "import keras\n",
    "from keras.datasets import reuters\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation, Dropout\n",
    "from keras.layers.noise import AlphaDropout\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "\n",
    "max_words = 1000\n",
    "batch_size = 32\n",
    "epochs = 25\n",
    "plot = True\n",
    "\n",
    "print('Loading data...')\n",
    "(x_train, y_train), (x_test, y_test) = reuters.load_data(num_words=max_words,\n",
    "                                                         test_split=0.2)\n",
    "print(len(x_train), 'train sequences')\n",
    "print(len(x_test), 'test sequences')\n",
    "\n",
    "num_classes = np.max(y_train) + 1\n",
    "print(num_classes, 'classes')\n",
    "\n",
    "print('Vectorizing sequence data...')\n",
    "tokenizer = Tokenizer(num_words=max_words)\n",
    "x_train = tokenizer.sequences_to_matrix(x_train, mode='binary')\n",
    "x_test = tokenizer.sequences_to_matrix(x_test, mode='binary')\n",
    "print('x_train shape:', x_train.shape)\n",
    "print('x_test shape:', x_test.shape)\n",
    "\n",
    "print('Convert class vector to binary class matrix '\n",
    "      '(for use with categorical_crossentropy)')\n",
    "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
    "y_test = keras.utils.to_categorical(y_test, num_classes)\n",
    "print('y_train shape:', y_train.shape)\n",
    "print('y_test shape:', y_test.shape)\n",
    "\n",
    "print('Building relu model...')\n",
    "model_relu = Sequential()\n",
    "model_relu.add(Dense(512, input_shape=(max_words,)))\n",
    "model_relu.add(Activation('relu'))\n",
    "model_relu.add(Dropout(0.5))\n",
    "model_relu.add(Dense(num_classes))\n",
    "model_relu.add(Activation('softmax'))\n",
    "\n",
    "model_relu.compile(loss='categorical_crossentropy',\n",
    "                   optimizer='adam',\n",
    "                   metrics=['accuracy'])\n",
    "\n",
    "history_relu = model_relu.fit(x_train, y_train,\n",
    "                              batch_size=batch_size,\n",
    "                              epochs=epochs,\n",
    "                              verbose=1,\n",
    "                              validation_split=0.1)\n",
    "\n",
    "\n",
    "score = model_relu.evaluate(x_test, y_test,\n",
    "                            batch_size=batch_size, verbose=1)\n",
    "\n",
    "print('RELU model results')\n",
    "print('Test score:', score[0])\n",
    "print('Test accuracy:', score[1])\n",
    "\n",
    "print('Building selu model...')\n",
    "model_selu = Sequential()\n",
    "model_selu.add(Dense(512, input_shape=(max_words,)))\n",
    "model_selu.add(Activation('selu'))\n",
    "model_selu.add(AlphaDropout(0.5))\n",
    "model_selu.add(Dense(num_classes))\n",
    "model_selu.add(Activation('softmax'))\n",
    "\n",
    "model_selu.compile(loss='categorical_crossentropy',\n",
    "                   optimizer='adam',\n",
    "                   metrics=['accuracy'])\n",
    "\n",
    "history_selu = model_selu.fit(x_train, y_train,\n",
    "                              batch_size=batch_size,\n",
    "                              epochs=epochs,\n",
    "                              verbose=1,\n",
    "                              validation_split=0.1)\n",
    "score = model_selu.evaluate(x_test, y_test,\n",
    "                            batch_size=batch_size, verbose=1)\n",
    "\n",
    "print('SELU model results')\n",
    "print('Test score:', score[0])\n",
    "print('Test accuracy:', score[1])\n",
    "\n",
    "if plot:\n",
    "    import matplotlib.pyplot as plt\n",
    "    plt.plot(range(epochs), history_relu.history['val_loss'], 'g-', label='RELU Val Loss')\n",
    "    plt.plot(range(epochs), history_selu.history['val_loss'], 'r-', label='SELU Val Loss')\n",
    "    plt.plot(range(epochs), history_relu.history['loss'], 'g--', label='RELU Loss')\n",
    "    plt.plot(range(epochs), history_selu.history['loss'], 'r--', label='SELU Loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    plt.legend()\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}