Model_03-Adam-k=10-amsgrad-00001.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Mandatory imports\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import log_loss\n",
    "from sklearn.model_selection import StratifiedKFold, StratifiedShuffleSplit\n",
    "from os.path import join as opj\n",
    "from matplotlib import pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import pylab\n",
    "plt.rcParams['figure.figsize'] = 10, 10\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train = pd.read_json(\"input/train.json\")\n",
    "target_train=train['is_iceberg']\n",
    "test = pd.read_json(\"input/test.json\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "target_train=train['is_iceberg']\n",
    "test['inc_angle']=pd.to_numeric(test['inc_angle'], errors='coerce')\n",
    "train['inc_angle']=pd.to_numeric(train['inc_angle'], errors='coerce')#We have only 133 NAs.\n",
    "train['inc_angle']=train['inc_angle'].fillna(method='pad')\n",
    "X_angle=train['inc_angle']\n",
    "test['inc_angle']=pd.to_numeric(test['inc_angle'], errors='coerce')\n",
    "X_test_angle=test['inc_angle']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Generate the training data\n",
    "X_band_1=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in train[\"band_1\"]])\n",
    "X_band_2=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in train[\"band_2\"]])\n",
    "X_band_3=(X_band_1+X_band_2)/2\n",
    "#X_band_3=np.array([np.full((75, 75), angel).astype(np.float32) for angel in train[\"inc_angle\"]])\n",
    "X_train = np.concatenate([X_band_1[:, :, :, np.newaxis]\n",
    "                          , X_band_2[:, :, :, np.newaxis]\n",
    "                         , X_band_3[:, :, :, np.newaxis]], axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_band_test_1=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in test[\"band_1\"]])\n",
    "X_band_test_2=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in test[\"band_2\"]])\n",
    "X_band_test_3=(X_band_test_1+X_band_test_2)/2\n",
    "#X_band_test_3=np.array([np.full((75, 75), angel).astype(np.float32) for angel in test[\"inc_angle\"]])\n",
    "X_test = np.concatenate([X_band_test_1[:, :, :, np.newaxis]\n",
    "                          , X_band_test_2[:, :, :, np.newaxis]\n",
    "                         , X_band_test_3[:, :, :, np.newaxis]], axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "#Import Keras.\n",
    "from matplotlib import pyplot\n",
    "from keras.optimizers import RMSprop\n",
    "# from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, MaxPooling2D, Dense, Dropout, Input, Flatten, Activation\n",
    "from keras.layers import GlobalMaxPooling2D\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "from keras.layers.merge import Concatenate\n",
    "from keras.models import Model\n",
    "from keras import initializers\n",
    "from keras.optimizers import Adam\n",
    "from keras.optimizers import rmsprop\n",
    "from keras.layers.advanced_activations import LeakyReLU, PReLU\n",
    "from keras.optimizers import SGD\n",
    "from keras.optimizers import Adadelta\n",
    "from keras.callbacks import ModelCheckpoint, Callback, EarlyStopping\n",
    "\n",
    "from keras.datasets import cifar10\n",
    "from keras.applications.inception_v3 import InceptionV3\n",
    "from keras.applications.vgg16 import VGG16\n",
    "from keras.applications.xception import Xception\n",
    "from keras.applications.mobilenet import MobileNet\n",
    "from keras.applications.vgg19 import VGG19\n",
    "from keras.layers import Concatenate, Dense, LSTM, Input, concatenate\n",
    "from keras.preprocessing import image\n",
    "from keras.applications.vgg16 import preprocess_input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Data Aug for multi-input\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "batch_size=64\n",
    "# Define the image transformations here\n",
    "gen = ImageDataGenerator(horizontal_flip = True,\n",
    "                         vertical_flip = True,\n",
    "                         width_shift_range = 0.,\n",
    "                         height_shift_range = 0.,\n",
    "                         channel_shift_range=0,\n",
    "                         zoom_range = 0.2,\n",
    "                         rotation_range = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Here is the function that merges our two generators\n",
    "# We use the exact same generator with the same random seed for both the y and angle arrays\n",
    "def gen_flow_for_two_inputs(X1, X2, y):\n",
    "    genX1 = gen.flow(X1,y,  batch_size=batch_size,seed=55)\n",
    "    genX2 = gen.flow(X1,X2, batch_size=batch_size,seed=55)\n",
    "    while True:\n",
    "            X1i = genX1.next()\n",
    "            X2i = genX2.next()\n",
    "            #Assert arrays are equal - this was for peace of mind, but slows down training\n",
    "            #np.testing.assert_array_equal(X1i[0],X2i[0])\n",
    "            yield [X1i[0], X2i[1]], X1i[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Finally create generator\n",
    "def get_callbacks(filepath, patience=2):\n",
    "    es = EarlyStopping('val_loss', patience=10, mode=\"min\")\n",
    "    msave = ModelCheckpoint(filepath, save_best_only=True)\n",
    "    return [es, msave]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def getVggAngleModel():\n",
    "    input_2 = Input(shape=[1], name=\"angle\")\n",
    "    angle_layer = Dense(1, )(input_2)\n",
    "    base_model = VGG16(weights='imagenet', include_top=False, \n",
    "                 input_shape=X_train.shape[1:], classes=1)\n",
    "    x = base_model.get_layer('block5_pool').output\n",
    "    \n",
    "\n",
    "    x = GlobalMaxPooling2D()(x)\n",
    "    merge_one = concatenate([x, angle_layer])\n",
    "    merge_one = Dense(512, activation='relu', name='fc2')(merge_one)\n",
    "    merge_one = Dropout(0.3)(merge_one)\n",
    "    merge_one = Dense(512, activation='relu', name='fc3')(merge_one)\n",
    "    merge_one = Dropout(0.3)(merge_one)\n",
    "    \n",
    "    predictions = Dense(1, activation='sigmoid')(merge_one)\n",
    "    \n",
    "    model = Model(inputs=[base_model.input, input_2], outputs=predictions)\n",
    "    \n",
    "    adam = Adam(lr=0.00001, beta_1=0.9, beta_2=0.999, epsilon=1e-8, decay=1e-6, amsgrad=True)\n",
    "    model.compile(loss='binary_crossentropy',\n",
    "                    optimizer=adam,\n",
    "                    metrics=['accuracy'])\n",
    "    \n",
    "#     sgd = SGD(lr=1e-3, decay=1e-6, momentum=0.9, nesterov=True)\n",
    "#     model.compile(loss='binary_crossentropy',\n",
    "#                   optimizer=sgd,\n",
    "#                   metrics=['accuracy'])\n",
    "    return model\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===================FOLD= 0\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 20s 840ms/step - loss: 0.7868 - acc: 0.5387 - val_loss: 0.5533 - val_acc: 0.7346\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 578ms/step - loss: 0.6298 - acc: 0.6449 - val_loss: 0.4546 - val_acc: 0.7840\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 580ms/step - loss: 0.5241 - acc: 0.7209 - val_loss: 0.3571 - val_acc: 0.8333\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 581ms/step - loss: 0.4109 - acc: 0.8038 - val_loss: 0.3131 - val_acc: 0.8519\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 14s 584ms/step - loss: 0.3926 - acc: 0.8134 - val_loss: 0.2890 - val_acc: 0.8642\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 12s 508ms/step - loss: 0.3272 - acc: 0.8539 - val_loss: 0.2917 - val_acc: 0.8642\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 14s 590ms/step - loss: 0.3167 - acc: 0.8660 - val_loss: 0.2873 - val_acc: 0.8642\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 14s 589ms/step - loss: 0.3177 - acc: 0.8711 - val_loss: 0.2424 - val_acc: 0.9074\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.2794 - acc: 0.8810 - val_loss: 0.2734 - val_acc: 0.8889\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2447 - acc: 0.8973 - val_loss: 0.2450 - val_acc: 0.9136\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2466 - acc: 0.8922 - val_loss: 0.2693 - val_acc: 0.8889\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2446 - acc: 0.8886 - val_loss: 0.2315 - val_acc: 0.9136\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2272 - acc: 0.8975 - val_loss: 0.2275 - val_acc: 0.9198\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2311 - acc: 0.9115 - val_loss: 0.2272 - val_acc: 0.9136\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2084 - acc: 0.9122 - val_loss: 0.2109 - val_acc: 0.9136\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2255 - acc: 0.9003 - val_loss: 0.2214 - val_acc: 0.9198\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1949 - acc: 0.9241 - val_loss: 0.2274 - val_acc: 0.9136\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2250 - acc: 0.9116 - val_loss: 0.2291 - val_acc: 0.9198\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1909 - acc: 0.9224 - val_loss: 0.2384 - val_acc: 0.9074\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2020 - acc: 0.9148 - val_loss: 0.2184 - val_acc: 0.9136\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2057 - acc: 0.9238 - val_loss: 0.2198 - val_acc: 0.9198\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1918 - acc: 0.9246 - val_loss: 0.2197 - val_acc: 0.9198\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 13s 527ms/step - loss: 0.1868 - acc: 0.9274 - val_loss: 0.2545 - val_acc: 0.9012\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1849 - acc: 0.9264 - val_loss: 0.2591 - val_acc: 0.9012\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1722 - acc: 0.9355 - val_loss: 0.2207 - val_acc: 0.9136\n",
      "Train loss: 0.173945174055\n",
      "Train accuracy: 0.928571428571\n",
      "Test loss: 0.210912205235\n",
      "Test accuracy: 0.913580246914\n",
      "\n",
      "===================FOLD= 1\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 17s 728ms/step - loss: 1.1261 - acc: 0.5239 - val_loss: 0.5270 - val_acc: 0.7143\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.9234 - acc: 0.6087 - val_loss: 0.4069 - val_acc: 0.8137\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.7651 - acc: 0.6623 - val_loss: 0.3728 - val_acc: 0.8261\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.6768 - acc: 0.7276 - val_loss: 0.3560 - val_acc: 0.8571\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.5703 - acc: 0.7945 - val_loss: 0.2692 - val_acc: 0.9006\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.5179 - acc: 0.7930 - val_loss: 0.2687 - val_acc: 0.8758\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.4233 - acc: 0.8339 - val_loss: 0.2822 - val_acc: 0.8820\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 14s 591ms/step - loss: 0.4192 - acc: 0.8420 - val_loss: 0.2398 - val_acc: 0.8944\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.3891 - acc: 0.8485 - val_loss: 0.2888 - val_acc: 0.8696\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.3738 - acc: 0.8547 - val_loss: 0.2675 - val_acc: 0.8944\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.4254 - acc: 0.8377 - val_loss: 0.3530 - val_acc: 0.8696\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.3383 - acc: 0.8701 - val_loss: 0.2446 - val_acc: 0.8944\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.3544 - acc: 0.8578 - val_loss: 0.2508 - val_acc: 0.8882\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 14s 590ms/step - loss: 0.3503 - acc: 0.8693 - val_loss: 0.1967 - val_acc: 0.9006\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.3268 - acc: 0.8780 - val_loss: 0.2583 - val_acc: 0.8882\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 510ms/step - loss: 0.3353 - acc: 0.8813 - val_loss: 0.1974 - val_acc: 0.9255\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2810 - acc: 0.8890 - val_loss: 0.2590 - val_acc: 0.8944\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2992 - acc: 0.8848 - val_loss: 0.2141 - val_acc: 0.9193\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2560 - acc: 0.9045 - val_loss: 0.2160 - val_acc: 0.9068\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.2868 - acc: 0.8886 - val_loss: 0.2115 - val_acc: 0.9193\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2512 - acc: 0.9102 - val_loss: 0.2043 - val_acc: 0.9130\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2638 - acc: 0.8969 - val_loss: 0.1963 - val_acc: 0.9130\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 13s 524ms/step - loss: 0.2495 - acc: 0.9012 - val_loss: 0.2039 - val_acc: 0.9130\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2429 - acc: 0.9096 - val_loss: 0.2498 - val_acc: 0.9068\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2532 - acc: 0.8972 - val_loss: 0.2007 - val_acc: 0.9317\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2362 - acc: 0.9046 - val_loss: 0.2140 - val_acc: 0.9379\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.2159 - acc: 0.9222 - val_loss: 0.2615 - val_acc: 0.9130\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2122 - acc: 0.9124 - val_loss: 0.2394 - val_acc: 0.9130\n",
      "Epoch 29/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2319 - acc: 0.9104 - val_loss: 0.2066 - val_acc: 0.9317\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2403 - acc: 0.9109 - val_loss: 0.1994 - val_acc: 0.9193\n",
      "Epoch 31/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2031 - acc: 0.9247 - val_loss: 0.2015 - val_acc: 0.9317\n",
      "Epoch 32/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1932 - acc: 0.9234 - val_loss: 0.2549 - val_acc: 0.9130\n",
      "Train loss: 0.153178880044\n",
      "Train accuracy: 0.939708939709\n",
      "Test loss: 0.19629342427\n",
      "Test accuracy: 0.913043478261\n",
      "\n",
      "===================FOLD= 2\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 16s 654ms/step - loss: 0.9112 - acc: 0.5608 - val_loss: 0.5239 - val_acc: 0.6894\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.7100 - acc: 0.6637 - val_loss: 0.3897 - val_acc: 0.8075\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.5886 - acc: 0.7526 - val_loss: 0.3412 - val_acc: 0.8137\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 591ms/step - loss: 0.5197 - acc: 0.7749 - val_loss: 0.3017 - val_acc: 0.8571\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.4377 - acc: 0.8169 - val_loss: 0.2960 - val_acc: 0.8509\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.4573 - acc: 0.8125 - val_loss: 0.2639 - val_acc: 0.8882\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.4077 - acc: 0.8306 - val_loss: 0.2316 - val_acc: 0.8758\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3867 - acc: 0.8405 - val_loss: 0.2700 - val_acc: 0.8758\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.3769 - acc: 0.8463 - val_loss: 0.2882 - val_acc: 0.8447\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3560 - acc: 0.8607 - val_loss: 0.2383 - val_acc: 0.8820\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.3601 - acc: 0.8591 - val_loss: 0.2274 - val_acc: 0.8882\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.3226 - acc: 0.8654 - val_loss: 0.2251 - val_acc: 0.9130\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2996 - acc: 0.8805 - val_loss: 0.2474 - val_acc: 0.8882\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 14s 590ms/step - loss: 0.2740 - acc: 0.8783 - val_loss: 0.2194 - val_acc: 0.8882\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2934 - acc: 0.8731 - val_loss: 0.2259 - val_acc: 0.9193\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.2760 - acc: 0.8911 - val_loss: 0.2201 - val_acc: 0.9317\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3247 - acc: 0.8810 - val_loss: 0.3038 - val_acc: 0.8509\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.3071 - acc: 0.8801 - val_loss: 0.2163 - val_acc: 0.9068\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2593 - acc: 0.8935 - val_loss: 0.2113 - val_acc: 0.9193\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2845 - acc: 0.8895 - val_loss: 0.2139 - val_acc: 0.9006\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2552 - acc: 0.8969 - val_loss: 0.2043 - val_acc: 0.9068\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2568 - acc: 0.8923 - val_loss: 0.2095 - val_acc: 0.9193\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 14s 587ms/step - loss: 0.2625 - acc: 0.8932 - val_loss: 0.1931 - val_acc: 0.9130\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2447 - acc: 0.9054 - val_loss: 0.2387 - val_acc: 0.9130\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2249 - acc: 0.9101 - val_loss: 0.2214 - val_acc: 0.9130\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.2397 - acc: 0.9015 - val_loss: 0.2348 - val_acc: 0.9130\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2293 - acc: 0.9027 - val_loss: 0.1946 - val_acc: 0.9317\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2236 - acc: 0.9055 - val_loss: 0.1872 - val_acc: 0.9317\n",
      "Epoch 29/100\n",
      "24/24 [==============================] - 14s 591ms/step - loss: 0.2054 - acc: 0.9167 - val_loss: 0.1863 - val_acc: 0.9255\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1856 - acc: 0.9260 - val_loss: 0.1949 - val_acc: 0.9317\n",
      "Epoch 31/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2064 - acc: 0.9173 - val_loss: 0.1965 - val_acc: 0.9130\n",
      "Epoch 32/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1969 - acc: 0.9149 - val_loss: 0.2067 - val_acc: 0.9193\n",
      "Epoch 33/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1901 - acc: 0.9270 - val_loss: 0.2286 - val_acc: 0.9193\n",
      "Epoch 34/100\n",
      "24/24 [==============================] - 12s 510ms/step - loss: 0.2122 - acc: 0.9182 - val_loss: 0.2012 - val_acc: 0.9317\n",
      "Epoch 35/100\n",
      "24/24 [==============================] - 12s 510ms/step - loss: 0.1798 - acc: 0.9196 - val_loss: 0.2309 - val_acc: 0.9193\n",
      "Epoch 36/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.1879 - acc: 0.9153 - val_loss: 0.1776 - val_acc: 0.9441\n",
      "Epoch 37/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1946 - acc: 0.9229 - val_loss: 0.2498 - val_acc: 0.8882\n",
      "Epoch 38/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1930 - acc: 0.9229 - val_loss: 0.2052 - val_acc: 0.9193\n",
      "Epoch 39/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1728 - acc: 0.9325 - val_loss: 0.1835 - val_acc: 0.9317\n",
      "Epoch 40/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1869 - acc: 0.9299 - val_loss: 0.1900 - val_acc: 0.9379\n",
      "Epoch 41/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1741 - acc: 0.9266 - val_loss: 0.1800 - val_acc: 0.9255\n",
      "Epoch 42/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1656 - acc: 0.9345 - val_loss: 0.1806 - val_acc: 0.9441\n",
      "Epoch 43/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1669 - acc: 0.9365 - val_loss: 0.2264 - val_acc: 0.9255\n",
      "Epoch 44/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1648 - acc: 0.9369 - val_loss: 0.2328 - val_acc: 0.9006\n",
      "Epoch 45/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1695 - acc: 0.9315 - val_loss: 0.2053 - val_acc: 0.9255\n",
      "Epoch 46/100\n",
      "24/24 [==============================] - 12s 505ms/step - loss: 0.1752 - acc: 0.9360 - val_loss: 0.2393 - val_acc: 0.8944\n",
      "Train loss: 0.10868784935\n",
      "Train accuracy: 0.95148995149\n",
      "Test loss: 0.177648523857\n",
      "Test accuracy: 0.944099378882\n",
      "\n",
      "===================FOLD= 3\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 17s 721ms/step - loss: 0.8187 - acc: 0.5674 - val_loss: 0.5889 - val_acc: 0.6562\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.6672 - acc: 0.6228 - val_loss: 0.4803 - val_acc: 0.7375\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.5672 - acc: 0.6821 - val_loss: 0.3700 - val_acc: 0.8500\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.4627 - acc: 0.7607 - val_loss: 0.3202 - val_acc: 0.8500\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.3575 - acc: 0.8329 - val_loss: 0.2849 - val_acc: 0.8875\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.3508 - acc: 0.8412 - val_loss: 0.2620 - val_acc: 0.8750\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.3352 - acc: 0.8554 - val_loss: 0.2653 - val_acc: 0.8875\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.3179 - acc: 0.8568 - val_loss: 0.2420 - val_acc: 0.9062\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2576 - acc: 0.8878 - val_loss: 0.2762 - val_acc: 0.9000\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2690 - acc: 0.8912 - val_loss: 0.2626 - val_acc: 0.9062\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2512 - acc: 0.8943 - val_loss: 0.2543 - val_acc: 0.8812\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2146 - acc: 0.9050 - val_loss: 0.2904 - val_acc: 0.9000\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2528 - acc: 0.8964 - val_loss: 0.2809 - val_acc: 0.8938\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 15s 630ms/step - loss: 0.2567 - acc: 0.8953 - val_loss: 0.2249 - val_acc: 0.9187\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.2517 - acc: 0.8938 - val_loss: 0.2420 - val_acc: 0.9062\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2548 - acc: 0.9015 - val_loss: 0.2291 - val_acc: 0.9125\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 14s 591ms/step - loss: 0.2095 - acc: 0.9122 - val_loss: 0.2246 - val_acc: 0.9125\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2165 - acc: 0.9101 - val_loss: 0.2840 - val_acc: 0.8875\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2262 - acc: 0.9021 - val_loss: 0.2253 - val_acc: 0.9187\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2028 - acc: 0.9145 - val_loss: 0.2548 - val_acc: 0.9062\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1900 - acc: 0.9241 - val_loss: 0.2365 - val_acc: 0.9125\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.1959 - acc: 0.9277 - val_loss: 0.2223 - val_acc: 0.9250\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 12s 520ms/step - loss: 0.1821 - acc: 0.9249 - val_loss: 0.2606 - val_acc: 0.9000\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1820 - acc: 0.9199 - val_loss: 0.3222 - val_acc: 0.8812\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2122 - acc: 0.9104 - val_loss: 0.2215 - val_acc: 0.9187\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1832 - acc: 0.9226 - val_loss: 0.2769 - val_acc: 0.9062\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1615 - acc: 0.9397 - val_loss: 0.2818 - val_acc: 0.8938\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1568 - acc: 0.9311 - val_loss: 0.2368 - val_acc: 0.9250\n",
      "Epoch 29/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1684 - acc: 0.9283 - val_loss: 0.2691 - val_acc: 0.8938\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1555 - acc: 0.9384 - val_loss: 0.3003 - val_acc: 0.9000\n",
      "Epoch 31/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1588 - acc: 0.9321 - val_loss: 0.2739 - val_acc: 0.9062\n",
      "Epoch 32/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1612 - acc: 0.9306 - val_loss: 0.2959 - val_acc: 0.9000\n",
      "Epoch 33/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1362 - acc: 0.9415 - val_loss: 0.2946 - val_acc: 0.9062\n",
      "Epoch 34/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1583 - acc: 0.9412 - val_loss: 0.2824 - val_acc: 0.9187\n",
      "Epoch 35/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1506 - acc: 0.9332 - val_loss: 0.2653 - val_acc: 0.9187\n",
      "Train loss: 0.119809226089\n",
      "Train accuracy: 0.953601108033\n",
      "Test loss: 0.221500104666\n",
      "Test accuracy: 0.91875\n",
      "\n",
      "===================FOLD= 4\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 16s 665ms/step - loss: 0.7497 - acc: 0.5456 - val_loss: 0.5660 - val_acc: 0.6687\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.5538 - acc: 0.6914 - val_loss: 0.3727 - val_acc: 0.8375\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.4374 - acc: 0.7779 - val_loss: 0.3284 - val_acc: 0.8625\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.3827 - acc: 0.8246 - val_loss: 0.2257 - val_acc: 0.9187\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3465 - acc: 0.8370 - val_loss: 0.2389 - val_acc: 0.9125\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.3190 - acc: 0.8562 - val_loss: 0.2043 - val_acc: 0.9313\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3008 - acc: 0.8598 - val_loss: 0.1986 - val_acc: 0.9500\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2791 - acc: 0.8638 - val_loss: 0.1899 - val_acc: 0.9437\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2748 - acc: 0.8723 - val_loss: 0.2216 - val_acc: 0.9187\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2841 - acc: 0.8695 - val_loss: 0.2021 - val_acc: 0.9250\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.2519 - acc: 0.8865 - val_loss: 0.2114 - val_acc: 0.9313\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2574 - acc: 0.8876 - val_loss: 0.1851 - val_acc: 0.9500\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2326 - acc: 0.8995 - val_loss: 0.1862 - val_acc: 0.9437\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2212 - acc: 0.8935 - val_loss: 0.1570 - val_acc: 0.9688\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2340 - acc: 0.8902 - val_loss: 0.1826 - val_acc: 0.9313\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 510ms/step - loss: 0.2382 - acc: 0.9070 - val_loss: 0.1634 - val_acc: 0.9563\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2134 - acc: 0.9049 - val_loss: 0.1769 - val_acc: 0.9563\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2132 - acc: 0.9073 - val_loss: 0.1576 - val_acc: 0.9625\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2241 - acc: 0.8992 - val_loss: 0.1746 - val_acc: 0.9500\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2010 - acc: 0.9140 - val_loss: 0.1717 - val_acc: 0.9437\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 14s 590ms/step - loss: 0.2072 - acc: 0.9124 - val_loss: 0.1489 - val_acc: 0.9625\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1984 - acc: 0.9179 - val_loss: 0.2427 - val_acc: 0.8938\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 12s 506ms/step - loss: 0.2115 - acc: 0.9071 - val_loss: 0.1541 - val_acc: 0.9688\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1815 - acc: 0.9209 - val_loss: 0.1889 - val_acc: 0.9375\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 14s 590ms/step - loss: 0.1957 - acc: 0.9213 - val_loss: 0.1481 - val_acc: 0.9688\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1759 - acc: 0.9239 - val_loss: 0.1781 - val_acc: 0.9313\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1962 - acc: 0.9174 - val_loss: 0.1815 - val_acc: 0.9250\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1614 - acc: 0.9301 - val_loss: 0.2242 - val_acc: 0.9250\n",
      "Epoch 29/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1765 - acc: 0.9282 - val_loss: 0.1837 - val_acc: 0.9187\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1479 - acc: 0.9376 - val_loss: 0.1657 - val_acc: 0.9625\n",
      "Epoch 31/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1751 - acc: 0.9223 - val_loss: 0.2223 - val_acc: 0.9000\n",
      "Epoch 32/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1535 - acc: 0.9407 - val_loss: 0.1643 - val_acc: 0.9437\n",
      "Epoch 33/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1761 - acc: 0.9235 - val_loss: 0.1593 - val_acc: 0.9625\n",
      "Epoch 34/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1610 - acc: 0.9378 - val_loss: 0.2664 - val_acc: 0.8938\n",
      "Epoch 35/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1729 - acc: 0.9246 - val_loss: 0.1718 - val_acc: 0.9437\n",
      "Train loss: 0.142912157644\n",
      "Train accuracy: 0.943213296399\n",
      "Test loss: 0.148136950284\n",
      "Test accuracy: 0.96875\n",
      "\n",
      "===================FOLD= 5\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 17s 688ms/step - loss: 0.7574 - acc: 0.5361 - val_loss: 0.5554 - val_acc: 0.7188\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.5744 - acc: 0.6759 - val_loss: 0.3696 - val_acc: 0.8438\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.4821 - acc: 0.7559 - val_loss: 0.3451 - val_acc: 0.8500\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.4177 - acc: 0.7909 - val_loss: 0.3616 - val_acc: 0.8250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 5/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3463 - acc: 0.8288 - val_loss: 0.2978 - val_acc: 0.8250\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3259 - acc: 0.8502 - val_loss: 0.2474 - val_acc: 0.8875\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2754 - acc: 0.8730 - val_loss: 0.2271 - val_acc: 0.8875\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.3288 - acc: 0.8606 - val_loss: 0.2263 - val_acc: 0.8875\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2825 - acc: 0.8643 - val_loss: 0.2583 - val_acc: 0.8750\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2640 - acc: 0.8806 - val_loss: 0.2634 - val_acc: 0.8812\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2655 - acc: 0.8813 - val_loss: 0.2846 - val_acc: 0.8812\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2643 - acc: 0.8819 - val_loss: 0.1934 - val_acc: 0.9062\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2306 - acc: 0.8992 - val_loss: 0.2293 - val_acc: 0.9062\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2406 - acc: 0.8969 - val_loss: 0.2100 - val_acc: 0.9125\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2256 - acc: 0.9028 - val_loss: 0.2243 - val_acc: 0.9000\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2212 - acc: 0.9037 - val_loss: 0.1901 - val_acc: 0.9125\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2316 - acc: 0.8940 - val_loss: 0.2695 - val_acc: 0.8875\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2301 - acc: 0.9064 - val_loss: 0.2281 - val_acc: 0.9125\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2087 - acc: 0.9101 - val_loss: 0.2566 - val_acc: 0.9000\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1803 - acc: 0.9161 - val_loss: 0.2252 - val_acc: 0.9187\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.2072 - acc: 0.9073 - val_loss: 0.2122 - val_acc: 0.9250\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1862 - acc: 0.9174 - val_loss: 0.2071 - val_acc: 0.9250\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 14s 587ms/step - loss: 0.2051 - acc: 0.9056 - val_loss: 0.1893 - val_acc: 0.9313\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2009 - acc: 0.9147 - val_loss: 0.2153 - val_acc: 0.9375\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.1893 - acc: 0.9213 - val_loss: 0.1790 - val_acc: 0.9375\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.1693 - acc: 0.9264 - val_loss: 0.1777 - val_acc: 0.9313\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1680 - acc: 0.9269 - val_loss: 0.1879 - val_acc: 0.9313\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1690 - acc: 0.9296 - val_loss: 0.2246 - val_acc: 0.9187\n",
      "Epoch 29/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1788 - acc: 0.9208 - val_loss: 0.1963 - val_acc: 0.9313\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1754 - acc: 0.9236 - val_loss: 0.2003 - val_acc: 0.9313\n",
      "Epoch 31/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.1607 - acc: 0.9314 - val_loss: 0.2218 - val_acc: 0.9375\n",
      "Epoch 32/100\n",
      "24/24 [==============================] - 12s 511ms/step - loss: 0.1477 - acc: 0.9428 - val_loss: 0.2215 - val_acc: 0.9250\n",
      "Epoch 33/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1798 - acc: 0.9197 - val_loss: 0.1817 - val_acc: 0.9500\n",
      "Epoch 34/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1594 - acc: 0.9410 - val_loss: 0.2221 - val_acc: 0.9313\n",
      "Epoch 35/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1371 - acc: 0.9417 - val_loss: 0.2231 - val_acc: 0.9313\n",
      "Epoch 36/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1426 - acc: 0.9399 - val_loss: 0.2147 - val_acc: 0.9250\n",
      "Train loss: 0.137870537091\n",
      "Train accuracy: 0.942520775623\n",
      "Test loss: 0.177746933699\n",
      "Test accuracy: 0.93125\n",
      "\n",
      "===================FOLD= 6\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 17s 693ms/step - loss: 0.6777 - acc: 0.5824 - val_loss: 0.5309 - val_acc: 0.6937\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.5126 - acc: 0.7236 - val_loss: 0.4380 - val_acc: 0.7625\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3942 - acc: 0.7982 - val_loss: 0.4632 - val_acc: 0.7875\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 596ms/step - loss: 0.3412 - acc: 0.8282 - val_loss: 0.4180 - val_acc: 0.8375\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3312 - acc: 0.8231 - val_loss: 0.3811 - val_acc: 0.8500\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2996 - acc: 0.8612 - val_loss: 0.3729 - val_acc: 0.8625\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2990 - acc: 0.8576 - val_loss: 0.3788 - val_acc: 0.8625\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.2805 - acc: 0.8761 - val_loss: 0.3608 - val_acc: 0.8750\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2550 - acc: 0.8862 - val_loss: 0.3551 - val_acc: 0.8562\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2361 - acc: 0.8943 - val_loss: 0.3439 - val_acc: 0.8750\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2257 - acc: 0.8919 - val_loss: 0.3634 - val_acc: 0.8625\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2344 - acc: 0.8966 - val_loss: 0.3509 - val_acc: 0.8812\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 15s 631ms/step - loss: 0.2444 - acc: 0.8920 - val_loss: 0.3355 - val_acc: 0.8750\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 12s 518ms/step - loss: 0.2289 - acc: 0.9037 - val_loss: 0.3452 - val_acc: 0.8875\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.2177 - acc: 0.9085 - val_loss: 0.3264 - val_acc: 0.8688\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2257 - acc: 0.8976 - val_loss: 0.3930 - val_acc: 0.8500\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2144 - acc: 0.9034 - val_loss: 0.3306 - val_acc: 0.8875\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2014 - acc: 0.9077 - val_loss: 0.3226 - val_acc: 0.8938\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1968 - acc: 0.9127 - val_loss: 0.3260 - val_acc: 0.8750\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1812 - acc: 0.9200 - val_loss: 0.3287 - val_acc: 0.8938\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1899 - acc: 0.9124 - val_loss: 0.3254 - val_acc: 0.8750\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1854 - acc: 0.9200 - val_loss: 0.3243 - val_acc: 0.8812\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 12s 508ms/step - loss: 0.1740 - acc: 0.9223 - val_loss: 0.3250 - val_acc: 0.8938\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1820 - acc: 0.9251 - val_loss: 0.3408 - val_acc: 0.8812\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.1710 - acc: 0.9256 - val_loss: 0.3121 - val_acc: 0.8875\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1774 - acc: 0.9262 - val_loss: 0.3197 - val_acc: 0.8812\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1738 - acc: 0.9261 - val_loss: 0.3181 - val_acc: 0.8688\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.1565 - acc: 0.9326 - val_loss: 0.3024 - val_acc: 0.9000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 29/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1574 - acc: 0.9332 - val_loss: 0.3255 - val_acc: 0.8938\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.1536 - acc: 0.9373 - val_loss: 0.3119 - val_acc: 0.8938\n",
      "Epoch 31/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1489 - acc: 0.9396 - val_loss: 0.3712 - val_acc: 0.8500\n",
      "Epoch 32/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1579 - acc: 0.9324 - val_loss: 0.3372 - val_acc: 0.8938\n",
      "Epoch 33/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1430 - acc: 0.9425 - val_loss: 0.3502 - val_acc: 0.8812\n",
      "Epoch 34/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1370 - acc: 0.9428 - val_loss: 0.3077 - val_acc: 0.8875\n",
      "Epoch 35/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1479 - acc: 0.9388 - val_loss: 0.3187 - val_acc: 0.8875\n",
      "Epoch 36/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.1281 - acc: 0.9471 - val_loss: 0.3217 - val_acc: 0.8875\n",
      "Epoch 37/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1567 - acc: 0.9348 - val_loss: 0.3524 - val_acc: 0.8750\n",
      "Epoch 38/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1349 - acc: 0.9471 - val_loss: 0.3697 - val_acc: 0.8875\n",
      "Train loss: 0.125817613155\n",
      "Train accuracy: 0.947368421053\n",
      "Test loss: 0.302441060543\n",
      "Test accuracy: 0.9\n",
      "\n",
      "===================FOLD= 7\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 17s 710ms/step - loss: 0.7491 - acc: 0.5604 - val_loss: 0.5571 - val_acc: 0.7375\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.6106 - acc: 0.6537 - val_loss: 0.4768 - val_acc: 0.7312\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.4911 - acc: 0.7386 - val_loss: 0.3565 - val_acc: 0.8313\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.4107 - acc: 0.7888 - val_loss: 0.3170 - val_acc: 0.8313\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3553 - acc: 0.8358 - val_loss: 0.2818 - val_acc: 0.8500\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3505 - acc: 0.8440 - val_loss: 0.2811 - val_acc: 0.8375\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.3004 - acc: 0.8625 - val_loss: 0.2565 - val_acc: 0.8625\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 12s 517ms/step - loss: 0.2926 - acc: 0.8709 - val_loss: 0.2625 - val_acc: 0.8625\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2975 - acc: 0.8599 - val_loss: 0.2386 - val_acc: 0.8750\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 14s 596ms/step - loss: 0.2536 - acc: 0.8868 - val_loss: 0.2337 - val_acc: 0.8875\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2590 - acc: 0.8839 - val_loss: 0.2421 - val_acc: 0.8812\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.2450 - acc: 0.8893 - val_loss: 0.2466 - val_acc: 0.8875\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.2347 - acc: 0.8989 - val_loss: 0.2669 - val_acc: 0.8688\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2449 - acc: 0.8972 - val_loss: 0.2157 - val_acc: 0.9062\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2118 - acc: 0.9078 - val_loss: 0.2317 - val_acc: 0.8875\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2421 - acc: 0.8982 - val_loss: 0.2113 - val_acc: 0.9187\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.1851 - acc: 0.9261 - val_loss: 0.2096 - val_acc: 0.9062\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2438 - acc: 0.8901 - val_loss: 0.2070 - val_acc: 0.9187\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1994 - acc: 0.9220 - val_loss: 0.2332 - val_acc: 0.8875\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2322 - acc: 0.8928 - val_loss: 0.1942 - val_acc: 0.9062\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2058 - acc: 0.9135 - val_loss: 0.2033 - val_acc: 0.9062\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.2033 - acc: 0.9148 - val_loss: 0.2242 - val_acc: 0.8875\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 12s 507ms/step - loss: 0.1963 - acc: 0.9188 - val_loss: 0.2095 - val_acc: 0.8938\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1914 - acc: 0.9207 - val_loss: 0.2182 - val_acc: 0.9125\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1937 - acc: 0.9163 - val_loss: 0.2090 - val_acc: 0.9125\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1858 - acc: 0.9225 - val_loss: 0.2180 - val_acc: 0.9125\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1606 - acc: 0.9353 - val_loss: 0.2308 - val_acc: 0.9000\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1849 - acc: 0.9290 - val_loss: 0.2078 - val_acc: 0.9062\n",
      "Epoch 29/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1649 - acc: 0.9378 - val_loss: 0.2243 - val_acc: 0.8875\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1827 - acc: 0.9208 - val_loss: 0.2304 - val_acc: 0.8875\n",
      "Train loss: 0.158932461715\n",
      "Train accuracy: 0.934210526316\n",
      "Test loss: 0.194206354022\n",
      "Test accuracy: 0.90625\n",
      "\n",
      "===================FOLD= 8\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 17s 722ms/step - loss: 0.7527 - acc: 0.5601 - val_loss: 0.6037 - val_acc: 0.6937\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.5742 - acc: 0.6795 - val_loss: 0.4480 - val_acc: 0.7750\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.4580 - acc: 0.7660 - val_loss: 0.3790 - val_acc: 0.7937\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3702 - acc: 0.8098 - val_loss: 0.3835 - val_acc: 0.7750\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3339 - acc: 0.8459 - val_loss: 0.3943 - val_acc: 0.7937\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2972 - acc: 0.8655 - val_loss: 0.3649 - val_acc: 0.8187\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2808 - acc: 0.8651 - val_loss: 0.3853 - val_acc: 0.8063\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2644 - acc: 0.8801 - val_loss: 0.3729 - val_acc: 0.8250\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2672 - acc: 0.8774 - val_loss: 0.3586 - val_acc: 0.8187\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2394 - acc: 0.8969 - val_loss: 0.3614 - val_acc: 0.8187\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2379 - acc: 0.9050 - val_loss: 0.3434 - val_acc: 0.8187\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2262 - acc: 0.9013 - val_loss: 0.3648 - val_acc: 0.8250\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2329 - acc: 0.9024 - val_loss: 0.3910 - val_acc: 0.8000\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2280 - acc: 0.9026 - val_loss: 0.3626 - val_acc: 0.8250\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2096 - acc: 0.9081 - val_loss: 0.3644 - val_acc: 0.8125\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2142 - acc: 0.9057 - val_loss: 0.3743 - val_acc: 0.8313\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2204 - acc: 0.9059 - val_loss: 0.3307 - val_acc: 0.8500\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2116 - acc: 0.9158 - val_loss: 0.3592 - val_acc: 0.8313\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.2037 - acc: 0.9119 - val_loss: 0.3788 - val_acc: 0.8375\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1935 - acc: 0.9187 - val_loss: 0.3689 - val_acc: 0.8187\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1949 - acc: 0.9244 - val_loss: 0.3516 - val_acc: 0.8438\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1794 - acc: 0.9339 - val_loss: 0.3711 - val_acc: 0.8313\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 12s 507ms/step - loss: 0.1734 - acc: 0.9260 - val_loss: 0.3538 - val_acc: 0.8313\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1947 - acc: 0.9156 - val_loss: 0.3742 - val_acc: 0.8375\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1831 - acc: 0.9225 - val_loss: 0.3697 - val_acc: 0.8500\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1646 - acc: 0.9366 - val_loss: 0.3860 - val_acc: 0.8375\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1791 - acc: 0.9256 - val_loss: 0.3899 - val_acc: 0.8375\n",
      "Train loss: 0.169736923108\n",
      "Train accuracy: 0.932132963989\n",
      "Test loss: 0.330704140663\n",
      "Test accuracy: 0.85\n",
      "\n",
      "===================FOLD= 9\n",
      "Epoch 1/100\n",
      "24/24 [==============================] - 18s 747ms/step - loss: 0.7674 - acc: 0.5788 - val_loss: 0.4980 - val_acc: 0.7688\n",
      "Epoch 2/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.5548 - acc: 0.7189 - val_loss: 0.3986 - val_acc: 0.8187\n",
      "Epoch 3/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.4188 - acc: 0.8088 - val_loss: 0.3011 - val_acc: 0.8438\n",
      "Epoch 4/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3459 - acc: 0.8503 - val_loss: 0.2606 - val_acc: 0.8938\n",
      "Epoch 5/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3775 - acc: 0.8396 - val_loss: 0.2730 - val_acc: 0.8875\n",
      "Epoch 6/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.3304 - acc: 0.8445 - val_loss: 0.2555 - val_acc: 0.8750\n",
      "Epoch 7/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.3138 - acc: 0.8695 - val_loss: 0.2403 - val_acc: 0.8812\n",
      "Epoch 8/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2879 - acc: 0.8757 - val_loss: 0.3493 - val_acc: 0.8250\n",
      "Epoch 9/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2737 - acc: 0.8906 - val_loss: 0.2312 - val_acc: 0.9000\n",
      "Epoch 10/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.2672 - acc: 0.8881 - val_loss: 0.2212 - val_acc: 0.8812\n",
      "Epoch 11/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.2446 - acc: 0.8936 - val_loss: 0.2483 - val_acc: 0.8625\n",
      "Epoch 12/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2255 - acc: 0.9033 - val_loss: 0.2415 - val_acc: 0.8938\n",
      "Epoch 13/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.3037 - acc: 0.8761 - val_loss: 0.2279 - val_acc: 0.8875\n",
      "Epoch 14/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.2406 - acc: 0.9080 - val_loss: 0.2083 - val_acc: 0.9062\n",
      "Epoch 15/100\n",
      "24/24 [==============================] - 14s 595ms/step - loss: 0.2418 - acc: 0.8832 - val_loss: 0.2065 - val_acc: 0.9125\n",
      "Epoch 16/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2096 - acc: 0.9137 - val_loss: 0.2112 - val_acc: 0.9000\n",
      "Epoch 17/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.2320 - acc: 0.9029 - val_loss: 0.2038 - val_acc: 0.9062\n",
      "Epoch 18/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.2092 - acc: 0.9091 - val_loss: 0.2178 - val_acc: 0.9062\n",
      "Epoch 19/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2277 - acc: 0.9140 - val_loss: 0.2116 - val_acc: 0.8938\n",
      "Epoch 20/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2135 - acc: 0.9143 - val_loss: 0.2052 - val_acc: 0.9062\n",
      "Epoch 21/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2388 - acc: 0.8954 - val_loss: 0.2373 - val_acc: 0.8938\n",
      "Epoch 22/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.2210 - acc: 0.9174 - val_loss: 0.2095 - val_acc: 0.9125\n",
      "Epoch 23/100\n",
      "24/24 [==============================] - 14s 588ms/step - loss: 0.1856 - acc: 0.9219 - val_loss: 0.1991 - val_acc: 0.9313\n",
      "Epoch 24/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.2039 - acc: 0.9158 - val_loss: 0.2064 - val_acc: 0.8938\n",
      "Epoch 25/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.1945 - acc: 0.9186 - val_loss: 0.1918 - val_acc: 0.9062\n",
      "Epoch 26/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1906 - acc: 0.9207 - val_loss: 0.2062 - val_acc: 0.9062\n",
      "Epoch 27/100\n",
      "24/24 [==============================] - 14s 594ms/step - loss: 0.1815 - acc: 0.9301 - val_loss: 0.1898 - val_acc: 0.9062\n",
      "Epoch 28/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1760 - acc: 0.9264 - val_loss: 0.1988 - val_acc: 0.9187\n",
      "Epoch 29/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.1766 - acc: 0.9257 - val_loss: 0.1969 - val_acc: 0.9125\n",
      "Epoch 30/100\n",
      "24/24 [==============================] - 14s 593ms/step - loss: 0.1566 - acc: 0.9365 - val_loss: 0.1894 - val_acc: 0.9250\n",
      "Epoch 31/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1616 - acc: 0.9360 - val_loss: 0.1933 - val_acc: 0.9062\n",
      "Epoch 32/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1832 - acc: 0.9261 - val_loss: 0.1904 - val_acc: 0.9062\n",
      "Epoch 33/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1765 - acc: 0.9308 - val_loss: 0.1976 - val_acc: 0.9062\n",
      "Epoch 34/100\n",
      "24/24 [==============================] - 12s 516ms/step - loss: 0.1672 - acc: 0.9249 - val_loss: 0.2003 - val_acc: 0.8938\n",
      "Epoch 35/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.1373 - acc: 0.9471 - val_loss: 0.1842 - val_acc: 0.9125\n",
      "Epoch 36/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1478 - acc: 0.9365 - val_loss: 0.1920 - val_acc: 0.8938\n",
      "Epoch 37/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1399 - acc: 0.9391 - val_loss: 0.1887 - val_acc: 0.9250\n",
      "Epoch 38/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1476 - acc: 0.9368 - val_loss: 0.1876 - val_acc: 0.9187\n",
      "Epoch 39/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1305 - acc: 0.9469 - val_loss: 0.1856 - val_acc: 0.9062\n",
      "Epoch 40/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1398 - acc: 0.9428 - val_loss: 0.2222 - val_acc: 0.9062\n",
      "Epoch 41/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1641 - acc: 0.9350 - val_loss: 0.2336 - val_acc: 0.9125\n",
      "Epoch 42/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.1288 - acc: 0.9498 - val_loss: 0.1868 - val_acc: 0.9125\n",
      "Epoch 43/100\n",
      "24/24 [==============================] - 14s 592ms/step - loss: 0.1459 - acc: 0.9406 - val_loss: 0.1806 - val_acc: 0.9187\n",
      "Epoch 44/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1195 - acc: 0.9524 - val_loss: 0.1934 - val_acc: 0.9187\n",
      "Epoch 45/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1287 - acc: 0.9487 - val_loss: 0.2088 - val_acc: 0.9062\n",
      "Epoch 46/100\n",
      "24/24 [==============================] - 12s 506ms/step - loss: 0.1260 - acc: 0.9509 - val_loss: 0.1866 - val_acc: 0.9063\n",
      "Epoch 47/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1132 - acc: 0.9578 - val_loss: 0.2161 - val_acc: 0.9125\n",
      "Epoch 48/100\n",
      "24/24 [==============================] - 12s 515ms/step - loss: 0.1353 - acc: 0.9430 - val_loss: 0.1953 - val_acc: 0.9125\n",
      "Epoch 49/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.0969 - acc: 0.9580 - val_loss: 0.2336 - val_acc: 0.8938\n",
      "Epoch 50/100\n",
      "24/24 [==============================] - 12s 510ms/step - loss: 0.1213 - acc: 0.9502 - val_loss: 0.2168 - val_acc: 0.9125\n",
      "Epoch 51/100\n",
      "24/24 [==============================] - 12s 512ms/step - loss: 0.1007 - acc: 0.9559 - val_loss: 0.2201 - val_acc: 0.9062\n",
      "Epoch 52/100\n",
      "24/24 [==============================] - 12s 513ms/step - loss: 0.1179 - acc: 0.9523 - val_loss: 0.2277 - val_acc: 0.9062\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 53/100\n",
      "24/24 [==============================] - 12s 514ms/step - loss: 0.0972 - acc: 0.9580 - val_loss: 0.2300 - val_acc: 0.9062\n",
      "Train loss: 0.0746470797327\n",
      "Train accuracy: 0.97783933518\n",
      "Test loss: 0.180586856604\n",
      "Test accuracy: 0.91875\n",
      "\n",
      " Train Log Loss Validation=  0.124790371608\n",
      " Test Log Loss Validation=  0.213980056827\n"
     ]
    }
   ],
   "source": [
    "#def myAngleCV(X_train, X_angle, X_test):\n",
    "K=10\n",
    "folds = list(StratifiedKFold(n_splits=K, shuffle=True, random_state=16).split(X_train, target_train))\n",
    "y_test_pred_log = 0\n",
    "y_train_pred_log=0\n",
    "y_valid_pred_log = 0.0*target_train\n",
    "for j, (train_idx, test_idx) in enumerate(folds):\n",
    "    print('\\n===================FOLD=',j)\n",
    "    X_train_cv = X_train[train_idx]\n",
    "    y_train_cv = target_train[train_idx]\n",
    "    X_holdout = X_train[test_idx]\n",
    "    Y_holdout= target_train[test_idx]\n",
    "    \n",
    "    #Angle\n",
    "    X_angle_cv=X_angle[train_idx]\n",
    "    X_angle_hold=X_angle[test_idx]\n",
    "    #define file path and get callbacks\n",
    "    file_path = \"%s_aug_model_weights.hdf5\"%j\n",
    "    callbacks = get_callbacks(filepath=file_path, patience=5)\n",
    "    gen_flow = gen_flow_for_two_inputs(X_train_cv, X_angle_cv, y_train_cv)\n",
    "    galaxyModel= getVggAngleModel()\n",
    "    histr = galaxyModel.fit_generator(\n",
    "            gen_flow,\n",
    "            steps_per_epoch=24,\n",
    "            epochs=100,\n",
    "            shuffle=True,\n",
    "            verbose=1,\n",
    "            validation_data=([X_holdout,X_angle_hold], Y_holdout),\n",
    "            callbacks=callbacks)\n",
    "     #Getting the Best Model\n",
    "    galaxyModel.load_weights(filepath=file_path)\n",
    "    #Getting Training Score\n",
    "    score = galaxyModel.evaluate([X_train_cv,X_angle_cv], y_train_cv, verbose=0)\n",
    "    print('Train loss:', score[0])\n",
    "    print('Train accuracy:', score[1])\n",
    "    #Getting Test Score\n",
    "    score = galaxyModel.evaluate([X_holdout,X_angle_hold], Y_holdout, verbose=0)\n",
    "    print('Test loss:', score[0])\n",
    "    print('Test accuracy:', score[1])\n",
    "\n",
    "    #Getting validation Score.\n",
    "    pred_valid=galaxyModel.predict([X_holdout,X_angle_hold])\n",
    "    y_valid_pred_log[test_idx] = pred_valid.reshape(pred_valid.shape[0])\n",
    "\n",
    "    #Getting Test Scores\n",
    "    temp_test=galaxyModel.predict([X_test, X_test_angle])\n",
    "    y_test_pred_log+=temp_test.reshape(temp_test.shape[0])\n",
    "\n",
    "    #Getting Train Scores\n",
    "    temp_train=galaxyModel.predict([X_train, X_angle])\n",
    "    y_train_pred_log+=temp_train.reshape(temp_train.shape[0])\n",
    "\n",
    "y_test_pred_log=y_test_pred_log/K\n",
    "y_train_pred_log=y_train_pred_log/K\n",
    "\n",
    "print('\\n Train Log Loss Validation= ',log_loss(target_train, y_train_pred_log))\n",
    "print(' Test Log Loss Validation= ',log_loss(target_train, y_valid_pred_log))\n",
    "#return y_test_pred_log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#preds=myAngleCV(X_train, X_angle, X_test)\n",
    "preds = y_test_pred_log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Submission to kaggle\n",
    "submission = pd.DataFrame()\n",
    "submission['id']=test['id']\n",
    "submission['is_iceberg']=preds\n",
    "submission.to_csv('sub_Adam_10folds_amsgrad_00001.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['val_loss', 'loss', 'acc', 'val_acc'])\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ntZIsIiKiVpJFRETU+gvIbcsJqqhrLgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f38a3850ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383fbc3da0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# list all data in history\n",
    "print(histr.history.keys())\n",
    "# summarize history for accuracy\n",
    "plt.plot(histr.history['acc'])\n",
    "plt.plot(histr.history['val_acc'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()\n",
    "# summarize history for loss\n",
    "plt.plot(histr.history['loss'])\n",
    "plt.plot(histr.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import matplotlib.image as mpimg\n",
    "\n",
    "min1 = np.min(X_holdout)\n",
    "max1 = np.max(X_holdout)\n",
    "diff1 = max1 - min1\n",
    "my_sum = lambda x: (x - min1)/diff1\n",
    "\n",
    "X_holdout2 = my_sum(X_holdout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13      1\n",
       "14      0\n",
       "38      1\n",
       "43      0\n",
       "44      0\n",
       "80      0\n",
       "104     1\n",
       "127     0\n",
       "137     0\n",
       "138     1\n",
       "141     1\n",
       "167     1\n",
       "195     1\n",
       "199     0\n",
       "207     1\n",
       "212     0\n",
       "213     1\n",
       "221     0\n",
       "240     0\n",
       "256     0\n",
       "267     1\n",
       "274     1\n",
       "276     1\n",
       "285     1\n",
       "290     0\n",
       "294     0\n",
       "300     1\n",
       "307     1\n",
       "311     1\n",
       "318     1\n",
       "       ..\n",
       "1363    1\n",
       "1364    0\n",
       "1373    1\n",
       "1380    1\n",
       "1394    0\n",
       "1411    0\n",
       "1435    0\n",
       "1440    1\n",
       "1445    1\n",
       "1447    1\n",
       "1460    1\n",
       "1470    0\n",
       "1488    0\n",
       "1495    0\n",
       "1504    1\n",
       "1516    0\n",
       "1525    0\n",
       "1533    0\n",
       "1537    0\n",
       "1539    0\n",
       "1541    0\n",
       "1542    0\n",
       "1549    0\n",
       "1552    0\n",
       "1559    0\n",
       "1566    0\n",
       "1567    0\n",
       "1577    0\n",
       "1590    0\n",
       "1602    0\n",
       "Name: is_iceberg, Length: 160, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_holdout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383cb23e80>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCMD/AGesuB89wUdgAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f38a3850ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[0], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[13])+ \",\"+\" iceberg_probability : \"+str(pred_valid[0])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383ca8d048>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUQAAAEICAYAAAAncI3RAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHCfgE6LjOE7AJ0THcZzAvwFQCqi3k7VGXwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383cae8c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[1], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[14])+ \",\"+\" iceberg_probability : \"+str(pred_valid[1])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383ca665c0>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383caedef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[2], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[38])+ \",\"+\" iceberg_probability : \"+str(pred_valid[2])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383c9c69e8>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUQAAAEICAYAAAAncI3RAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383ca9dac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[3], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[43])+ \",\"+\" iceberg_probability : \"+str(pred_valid[3])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383c9ab6a0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383c9fdf98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[6], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[104])+ \",\"+\" iceberg_probability : \"+str(pred_valid[6])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383c90ef98>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383c96a160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[9], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[138])+ \",\"+\" iceberg_probability : \"+str(pred_valid[9])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383c8f3898>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383c8c96a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[10], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[141])+ \",\"+\" iceberg_probability : \"+str(pred_valid[10])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383c858198>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIAAxKAZBEID/AVg0RmigpP+RAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383c8afba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[11], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[167])+ \",\"+\" iceberg_probability : \"+str(pred_valid[11])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383c83aa58>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383c80e8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[12], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[195])+ \",\"+\" iceberg_probability : \"+str(pred_valid[12])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f383c7a1358>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUUAAAEICAYAAADIsubvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nnAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f383c7efdd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_holdout2[13], cmap='gray', interpolation='nearest');\n",
    "tl = \"Actual label : \" +str(Y_holdout[199])+ \",\"+\" iceberg_probability : \"+str(pred_valid[13])\n",
    "plt.title(tl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}