SELU_Basic_CNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train shape: (60000, 28, 28, 1)\n",
      "60000 train samples\n",
      "10000 test samples\n",
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/200\n",
      "60000/60000 [==============================] - 6s - loss: 0.3118 - acc: 0.9070 - val_loss: 0.1044 - val_acc: 0.9689\n",
      "Epoch 2/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.1273 - acc: 0.9613 - val_loss: 0.0811 - val_acc: 0.9741\n",
      "Epoch 3/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0974 - acc: 0.9711 - val_loss: 0.0861 - val_acc: 0.9732\n",
      "Epoch 4/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0851 - acc: 0.9740 - val_loss: 0.0607 - val_acc: 0.9809\n",
      "Epoch 5/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0737 - acc: 0.9776 - val_loss: 0.0521 - val_acc: 0.9835\n",
      "Epoch 6/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0668 - acc: 0.9792 - val_loss: 0.0533 - val_acc: 0.9819\n",
      "Epoch 7/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0616 - acc: 0.9807 - val_loss: 0.0501 - val_acc: 0.9840\n",
      "Epoch 8/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0572 - acc: 0.9820 - val_loss: 0.0549 - val_acc: 0.9829\n",
      "Epoch 9/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0557 - acc: 0.9830 - val_loss: 0.0501 - val_acc: 0.9840\n",
      "Epoch 10/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0518 - acc: 0.9835 - val_loss: 0.0499 - val_acc: 0.9853\n",
      "Epoch 11/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0473 - acc: 0.9846 - val_loss: 0.0548 - val_acc: 0.9848\n",
      "Epoch 12/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0450 - acc: 0.9854 - val_loss: 0.0499 - val_acc: 0.9855\n",
      "Epoch 13/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0422 - acc: 0.9864 - val_loss: 0.0623 - val_acc: 0.9813\n",
      "Epoch 14/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0416 - acc: 0.9867 - val_loss: 0.0497 - val_acc: 0.9849\n",
      "Epoch 15/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0394 - acc: 0.9875 - val_loss: 0.0500 - val_acc: 0.9855\n",
      "Epoch 16/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0360 - acc: 0.9883 - val_loss: 0.0501 - val_acc: 0.9842\n",
      "Epoch 17/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0356 - acc: 0.9885 - val_loss: 0.0495 - val_acc: 0.9852\n",
      "Epoch 18/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0342 - acc: 0.9885 - val_loss: 0.0503 - val_acc: 0.9856\n",
      "Epoch 19/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0326 - acc: 0.9892 - val_loss: 0.0507 - val_acc: 0.9850\n",
      "Epoch 20/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0317 - acc: 0.9895 - val_loss: 0.0514 - val_acc: 0.9855\n",
      "Epoch 21/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0296 - acc: 0.9903 - val_loss: 0.0514 - val_acc: 0.9851\n",
      "Epoch 22/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0292 - acc: 0.9908 - val_loss: 0.0549 - val_acc: 0.9842\n",
      "Epoch 23/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0280 - acc: 0.9906 - val_loss: 0.0546 - val_acc: 0.9852\n",
      "Epoch 24/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0265 - acc: 0.9913 - val_loss: 0.0548 - val_acc: 0.9855\n",
      "Epoch 25/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0256 - acc: 0.9912 - val_loss: 0.0480 - val_acc: 0.9861\n",
      "Epoch 26/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0247 - acc: 0.9920 - val_loss: 0.0498 - val_acc: 0.9868\n",
      "Epoch 27/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0250 - acc: 0.9914 - val_loss: 0.0588 - val_acc: 0.9830\n",
      "Epoch 28/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0238 - acc: 0.9925 - val_loss: 0.0500 - val_acc: 0.9870\n",
      "Epoch 29/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0227 - acc: 0.9925 - val_loss: 0.0561 - val_acc: 0.9848\n",
      "Epoch 30/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0209 - acc: 0.9928 - val_loss: 0.0550 - val_acc: 0.9850\n",
      "Epoch 31/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0211 - acc: 0.9929 - val_loss: 0.0529 - val_acc: 0.9864\n",
      "Epoch 32/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0212 - acc: 0.9927 - val_loss: 0.0516 - val_acc: 0.9864\n",
      "Epoch 33/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0187 - acc: 0.9935 - val_loss: 0.0552 - val_acc: 0.9854\n",
      "Epoch 34/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0185 - acc: 0.9937 - val_loss: 0.0566 - val_acc: 0.9867\n",
      "Epoch 35/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0170 - acc: 0.9943 - val_loss: 0.0592 - val_acc: 0.9855\n",
      "Epoch 36/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0178 - acc: 0.9941 - val_loss: 0.0537 - val_acc: 0.9859\n",
      "Epoch 37/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0170 - acc: 0.9940 - val_loss: 0.0564 - val_acc: 0.9859\n",
      "Epoch 38/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0163 - acc: 0.9948 - val_loss: 0.0637 - val_acc: 0.9858\n",
      "Epoch 39/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0165 - acc: 0.9943 - val_loss: 0.0534 - val_acc: 0.9865\n",
      "Epoch 40/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0163 - acc: 0.9944 - val_loss: 0.0560 - val_acc: 0.9869\n",
      "Epoch 41/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0146 - acc: 0.9951 - val_loss: 0.0610 - val_acc: 0.9862\n",
      "Epoch 42/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0160 - acc: 0.9944 - val_loss: 0.0610 - val_acc: 0.9854\n",
      "Epoch 43/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0153 - acc: 0.9947 - val_loss: 0.0583 - val_acc: 0.9864\n",
      "Epoch 44/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0140 - acc: 0.9951 - val_loss: 0.0572 - val_acc: 0.9861\n",
      "Epoch 45/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0136 - acc: 0.9951 - val_loss: 0.0587 - val_acc: 0.9857\n",
      "Epoch 46/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0129 - acc: 0.9959 - val_loss: 0.0601 - val_acc: 0.9866\n",
      "Epoch 47/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0135 - acc: 0.9952 - val_loss: 0.0615 - val_acc: 0.9864\n",
      "Epoch 48/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0127 - acc: 0.9955 - val_loss: 0.0593 - val_acc: 0.9868\n",
      "Epoch 49/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0131 - acc: 0.9955 - val_loss: 0.0612 - val_acc: 0.9865\n",
      "Epoch 50/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0116 - acc: 0.9959 - val_loss: 0.0587 - val_acc: 0.9874\n",
      "Epoch 51/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0117 - acc: 0.9959 - val_loss: 0.0591 - val_acc: 0.9861\n",
      "Epoch 52/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0121 - acc: 0.9957 - val_loss: 0.0635 - val_acc: 0.9869\n",
      "Epoch 53/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0116 - acc: 0.9959 - val_loss: 0.0645 - val_acc: 0.9867\n",
      "Epoch 54/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0117 - acc: 0.9959 - val_loss: 0.0594 - val_acc: 0.9874\n",
      "Epoch 55/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0118 - acc: 0.9959 - val_loss: 0.0589 - val_acc: 0.9871\n",
      "Epoch 56/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0104 - acc: 0.9962 - val_loss: 0.0623 - val_acc: 0.9867\n",
      "Epoch 57/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0109 - acc: 0.9962 - val_loss: 0.0668 - val_acc: 0.9866\n",
      "Epoch 58/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0101 - acc: 0.9966 - val_loss: 0.0669 - val_acc: 0.9863\n",
      "Epoch 59/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0106 - acc: 0.9962 - val_loss: 0.0620 - val_acc: 0.9871\n",
      "Epoch 60/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0110 - acc: 0.9961 - val_loss: 0.0642 - val_acc: 0.9875\n",
      "Epoch 61/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0100 - acc: 0.9969 - val_loss: 0.0642 - val_acc: 0.9873\n",
      "Epoch 62/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0097 - acc: 0.9965 - val_loss: 0.0653 - val_acc: 0.9867\n",
      "Epoch 63/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 5s - loss: 0.0100 - acc: 0.9964 - val_loss: 0.0649 - val_acc: 0.9868\n",
      "Epoch 64/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0094 - acc: 0.9968 - val_loss: 0.0632 - val_acc: 0.9875\n",
      "Epoch 65/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0089 - acc: 0.9970 - val_loss: 0.0656 - val_acc: 0.9870\n",
      "Epoch 66/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0091 - acc: 0.9970 - val_loss: 0.0672 - val_acc: 0.9866\n",
      "Epoch 67/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0108 - acc: 0.9964 - val_loss: 0.0637 - val_acc: 0.9878\n",
      "Epoch 68/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0097 - acc: 0.9968 - val_loss: 0.0617 - val_acc: 0.9872\n",
      "Epoch 69/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0091 - acc: 0.9968 - val_loss: 0.0674 - val_acc: 0.9871\n",
      "Epoch 70/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0094 - acc: 0.9969 - val_loss: 0.0647 - val_acc: 0.9870\n",
      "Epoch 71/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0100 - acc: 0.9970 - val_loss: 0.0637 - val_acc: 0.9873\n",
      "Epoch 72/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0088 - acc: 0.9969 - val_loss: 0.0645 - val_acc: 0.9875\n",
      "Epoch 73/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0086 - acc: 0.9972 - val_loss: 0.0665 - val_acc: 0.9866\n",
      "Epoch 74/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0093 - acc: 0.9969 - val_loss: 0.0621 - val_acc: 0.9872\n",
      "Epoch 75/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0096 - acc: 0.9970 - val_loss: 0.0656 - val_acc: 0.9861\n",
      "Epoch 76/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0082 - acc: 0.9971 - val_loss: 0.0716 - val_acc: 0.9869\n",
      "Epoch 77/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0077 - acc: 0.9974 - val_loss: 0.0672 - val_acc: 0.9874\n",
      "Epoch 78/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0074 - acc: 0.9976 - val_loss: 0.0650 - val_acc: 0.9868\n",
      "Epoch 79/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0074 - acc: 0.9976 - val_loss: 0.0676 - val_acc: 0.9873\n",
      "Epoch 80/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0064 - acc: 0.9979 - val_loss: 0.0620 - val_acc: 0.9873\n",
      "Epoch 81/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0079 - acc: 0.9972 - val_loss: 0.0738 - val_acc: 0.9854\n",
      "Epoch 82/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0077 - acc: 0.9974 - val_loss: 0.0677 - val_acc: 0.9867\n",
      "Epoch 83/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0078 - acc: 0.9976 - val_loss: 0.0709 - val_acc: 0.9863\n",
      "Epoch 84/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0073 - acc: 0.9975 - val_loss: 0.0635 - val_acc: 0.9871\n",
      "Epoch 85/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0080 - acc: 0.9973 - val_loss: 0.0746 - val_acc: 0.9863\n",
      "Epoch 86/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0071 - acc: 0.9975 - val_loss: 0.0646 - val_acc: 0.9871\n",
      "Epoch 87/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0069 - acc: 0.9978 - val_loss: 0.0681 - val_acc: 0.9867\n",
      "Epoch 88/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0075 - acc: 0.9974 - val_loss: 0.0682 - val_acc: 0.9866\n",
      "Epoch 89/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0077 - acc: 0.9974 - val_loss: 0.0667 - val_acc: 0.9875\n",
      "Epoch 90/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0074 - acc: 0.9977 - val_loss: 0.0626 - val_acc: 0.9873\n",
      "Epoch 91/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0065 - acc: 0.9978 - val_loss: 0.0654 - val_acc: 0.9875\n",
      "Epoch 92/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0072 - acc: 0.9976 - val_loss: 0.0716 - val_acc: 0.9868\n",
      "Epoch 93/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0076 - acc: 0.9975 - val_loss: 0.0641 - val_acc: 0.9875\n",
      "Epoch 94/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0070 - acc: 0.9977 - val_loss: 0.0643 - val_acc: 0.9877\n",
      "Epoch 95/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0063 - acc: 0.9978 - val_loss: 0.0714 - val_acc: 0.9882\n",
      "Epoch 96/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0065 - acc: 0.9977 - val_loss: 0.0700 - val_acc: 0.9876\n",
      "Epoch 97/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0056 - acc: 0.9980 - val_loss: 0.0693 - val_acc: 0.9867\n",
      "Epoch 98/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0063 - acc: 0.9979 - val_loss: 0.0669 - val_acc: 0.9876\n",
      "Epoch 99/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0076 - acc: 0.9976 - val_loss: 0.0707 - val_acc: 0.9872\n",
      "Epoch 100/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0055 - acc: 0.9980 - val_loss: 0.0717 - val_acc: 0.9870\n",
      "Epoch 101/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0060 - acc: 0.9980 - val_loss: 0.0696 - val_acc: 0.9875\n",
      "Epoch 102/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0066 - acc: 0.9980 - val_loss: 0.0680 - val_acc: 0.9881\n",
      "Epoch 103/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0069 - acc: 0.9979 - val_loss: 0.0698 - val_acc: 0.9875\n",
      "Epoch 104/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0072 - acc: 0.9975 - val_loss: 0.0756 - val_acc: 0.9872\n",
      "Epoch 105/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0051 - acc: 0.9981 - val_loss: 0.0718 - val_acc: 0.9884\n",
      "Epoch 106/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0061 - acc: 0.9980 - val_loss: 0.0667 - val_acc: 0.9879\n",
      "Epoch 107/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0062 - acc: 0.9978 - val_loss: 0.0665 - val_acc: 0.9874\n",
      "Epoch 108/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0062 - acc: 0.9979 - val_loss: 0.0666 - val_acc: 0.9881\n",
      "Epoch 109/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0055 - acc: 0.9979 - val_loss: 0.0640 - val_acc: 0.9883\n",
      "Epoch 110/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0058 - acc: 0.9982 - val_loss: 0.0731 - val_acc: 0.9880\n",
      "Epoch 111/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0046 - acc: 0.9983 - val_loss: 0.0702 - val_acc: 0.9876\n",
      "Epoch 112/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0061 - acc: 0.9981 - val_loss: 0.0727 - val_acc: 0.9877\n",
      "Epoch 113/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0066 - acc: 0.9979 - val_loss: 0.0657 - val_acc: 0.9884\n",
      "Epoch 114/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0052 - acc: 0.9984 - val_loss: 0.0709 - val_acc: 0.9877\n",
      "Epoch 115/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0055 - acc: 0.9981 - val_loss: 0.0703 - val_acc: 0.9875\n",
      "Epoch 116/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0052 - acc: 0.9983 - val_loss: 0.0652 - val_acc: 0.9883\n",
      "Epoch 117/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0054 - acc: 0.9983 - val_loss: 0.0711 - val_acc: 0.9874\n",
      "Epoch 118/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0054 - acc: 0.9982 - val_loss: 0.0681 - val_acc: 0.9883\n",
      "Epoch 119/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0056 - acc: 0.9982 - val_loss: 0.0738 - val_acc: 0.9873\n",
      "Epoch 120/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0047 - acc: 0.9984 - val_loss: 0.0755 - val_acc: 0.9875\n",
      "Epoch 121/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0057 - acc: 0.9981 - val_loss: 0.0695 - val_acc: 0.9881\n",
      "Epoch 122/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0060 - acc: 0.9982 - val_loss: 0.0746 - val_acc: 0.9875\n",
      "Epoch 123/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0048 - acc: 0.9986 - val_loss: 0.0732 - val_acc: 0.9874\n",
      "Epoch 124/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0057 - acc: 0.9980 - val_loss: 0.0724 - val_acc: 0.9881\n",
      "Epoch 125/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0044 - acc: 0.9985 - val_loss: 0.0730 - val_acc: 0.9873\n",
      "Epoch 126/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 5s - loss: 0.0043 - acc: 0.9984 - val_loss: 0.0741 - val_acc: 0.9867\n",
      "Epoch 127/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0043 - acc: 0.9987 - val_loss: 0.0733 - val_acc: 0.9876\n",
      "Epoch 128/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0053 - acc: 0.9982 - val_loss: 0.0723 - val_acc: 0.9874\n",
      "Epoch 129/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0045 - acc: 0.9986 - val_loss: 0.0732 - val_acc: 0.9882\n",
      "Epoch 130/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0050 - acc: 0.9983 - val_loss: 0.0767 - val_acc: 0.9868\n",
      "Epoch 131/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0046 - acc: 0.9983 - val_loss: 0.0760 - val_acc: 0.9869\n",
      "Epoch 132/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0055 - acc: 0.9982 - val_loss: 0.0726 - val_acc: 0.9874\n",
      "Epoch 133/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0051 - acc: 0.9986 - val_loss: 0.0780 - val_acc: 0.9870\n",
      "Epoch 134/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0054 - acc: 0.9984 - val_loss: 0.0746 - val_acc: 0.9875\n",
      "Epoch 135/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0045 - acc: 0.9985 - val_loss: 0.0771 - val_acc: 0.9866\n",
      "Epoch 136/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0050 - acc: 0.9984 - val_loss: 0.0746 - val_acc: 0.9879\n",
      "Epoch 137/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0053 - acc: 0.9983 - val_loss: 0.0812 - val_acc: 0.9874\n",
      "Epoch 138/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0049 - acc: 0.9983 - val_loss: 0.0724 - val_acc: 0.9880\n",
      "Epoch 139/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0036 - acc: 0.9987 - val_loss: 0.0769 - val_acc: 0.9880\n",
      "Epoch 140/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0042 - acc: 0.9989 - val_loss: 0.0761 - val_acc: 0.9873\n",
      "Epoch 141/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0048 - acc: 0.9984 - val_loss: 0.0728 - val_acc: 0.9878\n",
      "Epoch 142/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0046 - acc: 0.9984 - val_loss: 0.0788 - val_acc: 0.9864\n",
      "Epoch 143/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0043 - acc: 0.9984 - val_loss: 0.0769 - val_acc: 0.9878\n",
      "Epoch 144/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9988 - val_loss: 0.0736 - val_acc: 0.9884\n",
      "Epoch 145/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0044 - acc: 0.9985 - val_loss: 0.0705 - val_acc: 0.9888\n",
      "Epoch 146/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0038 - acc: 0.9988 - val_loss: 0.0711 - val_acc: 0.9873\n",
      "Epoch 147/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0034 - acc: 0.9988 - val_loss: 0.0753 - val_acc: 0.9870\n",
      "Epoch 148/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0038 - acc: 0.9987 - val_loss: 0.0758 - val_acc: 0.9880\n",
      "Epoch 149/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9987 - val_loss: 0.0743 - val_acc: 0.9884\n",
      "Epoch 150/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0043 - acc: 0.9986 - val_loss: 0.0791 - val_acc: 0.9875\n",
      "Epoch 151/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0042 - acc: 0.9986 - val_loss: 0.0748 - val_acc: 0.9883\n",
      "Epoch 152/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0044 - acc: 0.9985 - val_loss: 0.0762 - val_acc: 0.9878\n",
      "Epoch 153/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0047 - acc: 0.9986 - val_loss: 0.0788 - val_acc: 0.9874\n",
      "Epoch 154/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0043 - acc: 0.9986 - val_loss: 0.0769 - val_acc: 0.9876\n",
      "Epoch 155/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0041 - acc: 0.9987 - val_loss: 0.0773 - val_acc: 0.9876\n",
      "Epoch 156/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0044 - acc: 0.9986 - val_loss: 0.0752 - val_acc: 0.9880\n",
      "Epoch 157/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9988 - val_loss: 0.0739 - val_acc: 0.9879\n",
      "Epoch 158/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9988 - val_loss: 0.0813 - val_acc: 0.9872\n",
      "Epoch 159/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0053 - acc: 0.9984 - val_loss: 0.0776 - val_acc: 0.9875\n",
      "Epoch 160/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9988 - val_loss: 0.0774 - val_acc: 0.9874\n",
      "Epoch 161/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9989 - val_loss: 0.0787 - val_acc: 0.9876\n",
      "Epoch 162/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9988 - val_loss: 0.0793 - val_acc: 0.9874\n",
      "Epoch 163/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0034 - acc: 0.9990 - val_loss: 0.0817 - val_acc: 0.9871\n",
      "Epoch 164/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0044 - acc: 0.9986 - val_loss: 0.0747 - val_acc: 0.9878\n",
      "Epoch 165/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0045 - acc: 0.9985 - val_loss: 0.0770 - val_acc: 0.9876\n",
      "Epoch 166/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0043 - acc: 0.9987 - val_loss: 0.0765 - val_acc: 0.9879\n",
      "Epoch 167/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0033 - acc: 0.9989 - val_loss: 0.0796 - val_acc: 0.9868\n",
      "Epoch 168/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9986 - val_loss: 0.0760 - val_acc: 0.9872\n",
      "Epoch 169/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0045 - acc: 0.9986 - val_loss: 0.0791 - val_acc: 0.9874\n",
      "Epoch 170/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0038 - acc: 0.9987 - val_loss: 0.0721 - val_acc: 0.9881\n",
      "Epoch 171/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9989 - val_loss: 0.0767 - val_acc: 0.9884\n",
      "Epoch 172/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0034 - acc: 0.9990 - val_loss: 0.0785 - val_acc: 0.9883\n",
      "Epoch 173/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0046 - acc: 0.9986 - val_loss: 0.0739 - val_acc: 0.9885\n",
      "Epoch 174/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0036 - acc: 0.9988 - val_loss: 0.0696 - val_acc: 0.9880\n",
      "Epoch 175/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0041 - acc: 0.9987 - val_loss: 0.0722 - val_acc: 0.9890\n",
      "Epoch 176/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0036 - acc: 0.9989 - val_loss: 0.0692 - val_acc: 0.9886\n",
      "Epoch 177/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0034 - acc: 0.9988 - val_loss: 0.0788 - val_acc: 0.9875\n",
      "Epoch 178/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0040 - acc: 0.9989 - val_loss: 0.0745 - val_acc: 0.9887\n",
      "Epoch 179/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0035 - acc: 0.9988 - val_loss: 0.0782 - val_acc: 0.9878\n",
      "Epoch 180/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9986 - val_loss: 0.0788 - val_acc: 0.9879\n",
      "Epoch 181/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9986 - val_loss: 0.0748 - val_acc: 0.9885\n",
      "Epoch 182/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0049 - acc: 0.9985 - val_loss: 0.0743 - val_acc: 0.9879\n",
      "Epoch 183/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0032 - acc: 0.9989 - val_loss: 0.0755 - val_acc: 0.9880\n",
      "Epoch 184/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9988 - val_loss: 0.0756 - val_acc: 0.9880\n",
      "Epoch 185/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0035 - acc: 0.9989 - val_loss: 0.0773 - val_acc: 0.9881\n",
      "Epoch 186/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9988 - val_loss: 0.0793 - val_acc: 0.9881\n",
      "Epoch 187/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0036 - acc: 0.9989 - val_loss: 0.0793 - val_acc: 0.9879\n",
      "Epoch 188/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0034 - acc: 0.9990 - val_loss: 0.0746 - val_acc: 0.9882\n",
      "Epoch 189/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 5s - loss: 0.0036 - acc: 0.9989 - val_loss: 0.0782 - val_acc: 0.9881\n",
      "Epoch 190/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9988 - val_loss: 0.0809 - val_acc: 0.9878\n",
      "Epoch 191/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0041 - acc: 0.9988 - val_loss: 0.0789 - val_acc: 0.9886\n",
      "Epoch 192/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0037 - acc: 0.9989 - val_loss: 0.0774 - val_acc: 0.9875\n",
      "Epoch 193/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0039 - acc: 0.9987 - val_loss: 0.0747 - val_acc: 0.9883\n",
      "Epoch 194/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0035 - acc: 0.9987 - val_loss: 0.0784 - val_acc: 0.9873\n",
      "Epoch 195/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0040 - acc: 0.9989 - val_loss: 0.0768 - val_acc: 0.9883\n",
      "Epoch 196/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0041 - acc: 0.9986 - val_loss: 0.0807 - val_acc: 0.9884\n",
      "Epoch 197/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0035 - acc: 0.9990 - val_loss: 0.0742 - val_acc: 0.9885\n",
      "Epoch 198/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0038 - acc: 0.9987 - val_loss: 0.0817 - val_acc: 0.9876\n",
      "Epoch 199/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0044 - acc: 0.9986 - val_loss: 0.0776 - val_acc: 0.9879\n",
      "Epoch 200/200\n",
      "60000/60000 [==============================] - 5s - loss: 0.0034 - acc: 0.9988 - val_loss: 0.0825 - val_acc: 0.9880\n",
      "Test loss: 0.0824557533961\n",
      "Test accuracy: 0.988\n"
     ]
    }
   ],
   "source": [
    "'''Trains a simple convnet on the MNIST dataset.\n",
    "\n",
    "NOTE: This network does not use AlphaDropout expressed in SELU paper.\n",
    "\n",
    "Gets to 99.25% test accuracy after 28 epochs\n",
    "'''\n",
    "\n",
    "from __future__ import print_function\n",
    "import keras\n",
    "from keras.datasets import mnist\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras import backend as K\n",
    "\n",
    "batch_size = 128\n",
    "num_classes = 10\n",
    "epochs = 200\n",
    "\n",
    "# input image dimensions\n",
    "img_rows, img_cols = 28, 28\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",
    "if K.image_data_format() == 'channels_first':\n",
    "    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)\n",
    "    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)\n",
    "    input_shape = (1, img_rows, img_cols)\n",
    "else:\n",
    "    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)\n",
    "    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)\n",
    "    input_shape = (img_rows, img_cols, 1)\n",
    "\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:', x_train.shape)\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(Conv2D(32, kernel_size=(3, 3),\n",
    "                 activation='selu',\n",
    "                 input_shape=input_shape))\n",
    "model.add(Conv2D(64, (3, 3), activation='selu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(128, activation='selu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(num_classes, activation='softmax'))\n",
    "\n",
    "model.compile(loss=keras.losses.categorical_crossentropy,\n",
    "              optimizer=keras.optimizers.Adadelta(),\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "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",
    "score = model.evaluate(x_test, y_test, verbose=0)\n",
    "print('Test loss:', score[0])\n",
    "print('Test accuracy:', score[1])"
   ]
  },
  {
   "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
}