chap3-1_binary_classification_IMDB.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"chap3-1_binary_classification_IMDB.ipynb","provenance":[],"collapsed_sections":[]},"kernelspec":{"name":"python3","display_name":"Python 3"}},"cells":[{"cell_type":"markdown","metadata":{"id":"ugISptKTce0b","colab_type":"text"},"source":["# Binary classification"]},{"cell_type":"markdown","metadata":{"id":"86zSbKyOWhAi","colab_type":"text"},"source":["<strong>Abstract</strong>  \n","In this notebook, neural networks will be build for classifying movie reviews provided by IMDB. This is a <strong>binary classification</strong> example.\n","\n","<strong>Reference</strong>  \n","See pages 68-77 of \"<strong>Deep Learning with Python</strong>\" by Francois Chollet (2018). \n","\n","<strong>Summary (page 77)</strong>  \n","- For a binary classification problem, the final activation function should be sigmoid, which returns a value between 0 and 1.\n","- An appropreate loss function is binary_crossentropy.\n","- Check whether there is a problem of overfitting.\n"]},{"cell_type":"code","metadata":{"id":"qIYhnTK1UMwv","colab_type":"code","colab":{}},"source":["import numpy as np\n","import matplotlib.pyplot as plt\n","\n","import warnings\n","warnings.filterwarnings(\"ignore\")"],"execution_count":0,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"O5LpNOUQVWBV","colab_type":"text"},"source":["## Loading IMDB dataset"]},{"cell_type":"markdown","metadata":{"id":"ntQNq1lZVa8L","colab_type":"text"},"source":["The IMDB dataset will be loaded. This dataset is already preprocessed. The sequences of words are turned into sequences of integers. Each integer stands for a specific word. The label data (train_labels and test_labels) are lists of 0s and 1s, where 0 and 1 are for negative and positive reviews, respectively. "]},{"cell_type":"code","metadata":{"id":"Rx9z47rpUARO","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":136},"outputId":"d5bea7e1-6403-4188-ebe3-81f898e59c8a","executionInfo":{"status":"ok","timestamp":1577080916442,"user_tz":-60,"elapsed":6602,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["from keras.datasets import imdb\n","\n","(train_data, train_labels), (test_data, test_labels) = imdb.load_data(num_words = 10000)\n","\n","print(\"train_data shape: {}\".format(train_data.shape))\n","print(\"train_labels shape: {}\".format(train_labels.shape))\n","print(\"test_data shape: {}\".format(test_data.shape))\n","print(\"test_labels shape: {}\".format(test_labels.shape))\n"],"execution_count":14,"outputs":[{"output_type":"stream","text":["train_data shape: (25000,)\n","train_labels shape: (25000,)\n","test_data shape: (25000,)\n","test_labels shape: (25000,)\n","-----------------------\n","\n","The number of words in the first training data:  218\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"r8dj--tuUNKW","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":173},"outputId":"c8289206-0aaf-448b-9529-4dd10dfa460c","executionInfo":{"status":"ok","timestamp":1577080650340,"user_tz":-60,"elapsed":783,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["print(\"the first 5 train_data: \\n{}\".format(train_data[:5]))\n","print(\"the first 5 train_labes: \\n{}\".format(train_labels[:5]))"],"execution_count":13,"outputs":[{"output_type":"stream","text":["the first 5 train_data: \n","[list([1, 14, 22, 16, 43, 530, 973, 1622, 1385, 65, 458, 4468, 66, 3941, 4, 173, 36, 256, 5, 25, 100, 43, 838, 112, 50, 670, 2, 9, 35, 480, 284, 5, 150, 4, 172, 112, 167, 2, 336, 385, 39, 4, 172, 4536, 1111, 17, 546, 38, 13, 447, 4, 192, 50, 16, 6, 147, 2025, 19, 14, 22, 4, 1920, 4613, 469, 4, 22, 71, 87, 12, 16, 43, 530, 38, 76, 15, 13, 1247, 4, 22, 17, 515, 17, 12, 16, 626, 18, 2, 5, 62, 386, 12, 8, 316, 8, 106, 5, 4, 2223, 5244, 16, 480, 66, 3785, 33, 4, 130, 12, 16, 38, 619, 5, 25, 124, 51, 36, 135, 48, 25, 1415, 33, 6, 22, 12, 215, 28, 77, 52, 5, 14, 407, 16, 82, 2, 8, 4, 107, 117, 5952, 15, 256, 4, 2, 7, 3766, 5, 723, 36, 71, 43, 530, 476, 26, 400, 317, 46, 7, 4, 2, 1029, 13, 104, 88, 4, 381, 15, 297, 98, 32, 2071, 56, 26, 141, 6, 194, 7486, 18, 4, 226, 22, 21, 134, 476, 26, 480, 5, 144, 30, 5535, 18, 51, 36, 28, 224, 92, 25, 104, 4, 226, 65, 16, 38, 1334, 88, 12, 16, 283, 5, 16, 4472, 113, 103, 32, 15, 16, 5345, 19, 178, 32])\n"," list([1, 194, 1153, 194, 8255, 78, 228, 5, 6, 1463, 4369, 5012, 134, 26, 4, 715, 8, 118, 1634, 14, 394, 20, 13, 119, 954, 189, 102, 5, 207, 110, 3103, 21, 14, 69, 188, 8, 30, 23, 7, 4, 249, 126, 93, 4, 114, 9, 2300, 1523, 5, 647, 4, 116, 9, 35, 8163, 4, 229, 9, 340, 1322, 4, 118, 9, 4, 130, 4901, 19, 4, 1002, 5, 89, 29, 952, 46, 37, 4, 455, 9, 45, 43, 38, 1543, 1905, 398, 4, 1649, 26, 6853, 5, 163, 11, 3215, 2, 4, 1153, 9, 194, 775, 7, 8255, 2, 349, 2637, 148, 605, 2, 8003, 15, 123, 125, 68, 2, 6853, 15, 349, 165, 4362, 98, 5, 4, 228, 9, 43, 2, 1157, 15, 299, 120, 5, 120, 174, 11, 220, 175, 136, 50, 9, 4373, 228, 8255, 5, 2, 656, 245, 2350, 5, 4, 9837, 131, 152, 491, 18, 2, 32, 7464, 1212, 14, 9, 6, 371, 78, 22, 625, 64, 1382, 9, 8, 168, 145, 23, 4, 1690, 15, 16, 4, 1355, 5, 28, 6, 52, 154, 462, 33, 89, 78, 285, 16, 145, 95])\n"," list([1, 14, 47, 8, 30, 31, 7, 4, 249, 108, 7, 4, 5974, 54, 61, 369, 13, 71, 149, 14, 22, 112, 4, 2401, 311, 12, 16, 3711, 33, 75, 43, 1829, 296, 4, 86, 320, 35, 534, 19, 263, 4821, 1301, 4, 1873, 33, 89, 78, 12, 66, 16, 4, 360, 7, 4, 58, 316, 334, 11, 4, 1716, 43, 645, 662, 8, 257, 85, 1200, 42, 1228, 2578, 83, 68, 3912, 15, 36, 165, 1539, 278, 36, 69, 2, 780, 8, 106, 14, 6905, 1338, 18, 6, 22, 12, 215, 28, 610, 40, 6, 87, 326, 23, 2300, 21, 23, 22, 12, 272, 40, 57, 31, 11, 4, 22, 47, 6, 2307, 51, 9, 170, 23, 595, 116, 595, 1352, 13, 191, 79, 638, 89, 2, 14, 9, 8, 106, 607, 624, 35, 534, 6, 227, 7, 129, 113])\n"," list([1, 4, 2, 2, 33, 2804, 4, 2040, 432, 111, 153, 103, 4, 1494, 13, 70, 131, 67, 11, 61, 2, 744, 35, 3715, 761, 61, 5766, 452, 9214, 4, 985, 7, 2, 59, 166, 4, 105, 216, 1239, 41, 1797, 9, 15, 7, 35, 744, 2413, 31, 8, 4, 687, 23, 4, 2, 7339, 6, 3693, 42, 38, 39, 121, 59, 456, 10, 10, 7, 265, 12, 575, 111, 153, 159, 59, 16, 1447, 21, 25, 586, 482, 39, 4, 96, 59, 716, 12, 4, 172, 65, 9, 579, 11, 6004, 4, 1615, 5, 2, 7, 5168, 17, 13, 7064, 12, 19, 6, 464, 31, 314, 11, 2, 6, 719, 605, 11, 8, 202, 27, 310, 4, 3772, 3501, 8, 2722, 58, 10, 10, 537, 2116, 180, 40, 14, 413, 173, 7, 263, 112, 37, 152, 377, 4, 537, 263, 846, 579, 178, 54, 75, 71, 476, 36, 413, 263, 2504, 182, 5, 17, 75, 2306, 922, 36, 279, 131, 2895, 17, 2867, 42, 17, 35, 921, 2, 192, 5, 1219, 3890, 19, 2, 217, 4122, 1710, 537, 2, 1236, 5, 736, 10, 10, 61, 403, 9, 2, 40, 61, 4494, 5, 27, 4494, 159, 90, 263, 2311, 4319, 309, 8, 178, 5, 82, 4319, 4, 65, 15, 9225, 145, 143, 5122, 12, 7039, 537, 746, 537, 537, 15, 7979, 4, 2, 594, 7, 5168, 94, 9096, 3987, 2, 11, 2, 4, 538, 7, 1795, 246, 2, 9, 2, 11, 635, 14, 9, 51, 408, 12, 94, 318, 1382, 12, 47, 6, 2683, 936, 5, 6307, 2, 19, 49, 7, 4, 1885, 2, 1118, 25, 80, 126, 842, 10, 10, 2, 2, 4726, 27, 4494, 11, 1550, 3633, 159, 27, 341, 29, 2733, 19, 4185, 173, 7, 90, 2, 8, 30, 11, 4, 1784, 86, 1117, 8, 3261, 46, 11, 2, 21, 29, 9, 2841, 23, 4, 1010, 2, 793, 6, 2, 1386, 1830, 10, 10, 246, 50, 9, 6, 2750, 1944, 746, 90, 29, 2, 8, 124, 4, 882, 4, 882, 496, 27, 2, 2213, 537, 121, 127, 1219, 130, 5, 29, 494, 8, 124, 4, 882, 496, 4, 341, 7, 27, 846, 10, 10, 29, 9, 1906, 8, 97, 6, 236, 2, 1311, 8, 4, 2, 7, 31, 7, 2, 91, 2, 3987, 70, 4, 882, 30, 579, 42, 9, 12, 32, 11, 537, 10, 10, 11, 14, 65, 44, 537, 75, 2, 1775, 3353, 2, 1846, 4, 2, 7, 154, 5, 4, 518, 53, 2, 2, 7, 3211, 882, 11, 399, 38, 75, 257, 3807, 19, 2, 17, 29, 456, 4, 65, 7, 27, 205, 113, 10, 10, 2, 4, 2, 2, 9, 242, 4, 91, 1202, 2, 5, 2070, 307, 22, 7, 5168, 126, 93, 40, 2, 13, 188, 1076, 3222, 19, 4, 2, 7, 2348, 537, 23, 53, 537, 21, 82, 40, 2, 13, 2, 14, 280, 13, 219, 4, 2, 431, 758, 859, 4, 953, 1052, 2, 7, 5991, 5, 94, 40, 25, 238, 60, 2, 4, 2, 804, 2, 7, 4, 9941, 132, 8, 67, 6, 22, 15, 9, 283, 8, 5168, 14, 31, 9, 242, 955, 48, 25, 279, 2, 23, 12, 1685, 195, 25, 238, 60, 796, 2, 4, 671, 7, 2804, 5, 4, 559, 154, 888, 7, 726, 50, 26, 49, 7008, 15, 566, 30, 579, 21, 64, 2574])\n"," list([1, 249, 1323, 7, 61, 113, 10, 10, 13, 1637, 14, 20, 56, 33, 2401, 18, 457, 88, 13, 2626, 1400, 45, 3171, 13, 70, 79, 49, 706, 919, 13, 16, 355, 340, 355, 1696, 96, 143, 4, 22, 32, 289, 7, 61, 369, 71, 2359, 5, 13, 16, 131, 2073, 249, 114, 249, 229, 249, 20, 13, 28, 126, 110, 13, 473, 8, 569, 61, 419, 56, 429, 6, 1513, 18, 35, 534, 95, 474, 570, 5, 25, 124, 138, 88, 12, 421, 1543, 52, 725, 6397, 61, 419, 11, 13, 1571, 15, 1543, 20, 11, 4, 2, 5, 296, 12, 3524, 5, 15, 421, 128, 74, 233, 334, 207, 126, 224, 12, 562, 298, 2167, 1272, 7, 2601, 5, 516, 988, 43, 8, 79, 120, 15, 595, 13, 784, 25, 3171, 18, 165, 170, 143, 19, 14, 5, 7224, 6, 226, 251, 7, 61, 113])]\n","the first 5 train_labes: \n","[1 0 0 1 0]\n"],"name":"stdout"}]},{"cell_type":"markdown","metadata":{"id":"r6JuLi3AYW8V","colab_type":"text"},"source":["Let us perform 0 paddding so that all data have the same length."]},{"cell_type":"code","metadata":{"id":"cer22S6DXmOO","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":119},"outputId":"07701e01-bfc7-4264-e922-a3b61fee51c2","executionInfo":{"status":"ok","timestamp":1577081622978,"user_tz":-60,"elapsed":2917,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["def vectorize_sequences(sequences, dimension=10000):\n","  results = np.zeros((len(sequences),dimension))\n","  for i , sequence in enumerate(sequences):\n","    results[i,sequence] = 1.\n","  \n","  return results\n","\n","x_train = vectorize_sequences(train_data)\n","x_test = vectorize_sequences(test_data)\n","\n","print(\"length of the first training data: \", len(x_train[0]))\n","print(\"length of the second training data: \", len(x_train[1]))\n","print(\"length of the third training data: \", len(x_train[2]))\n","\n","print(\"length of the first test data: \", len(x_test[0]))\n","print(\"length of the second test data: \", len(x_test[1]))\n","print(\"length of the third test data: \", len(x_test[2]))"],"execution_count":22,"outputs":[{"output_type":"stream","text":["length of the first training data:  10000\n","length of the second training data:  10000\n","length of the third training data:  10000\n","length of the first test data:  10000\n","length of the second test data:  10000\n","length of the third test data:  10000\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"_bYKwfWgZGZt","colab_type":"code","colab":{}},"source":["y_train = np.asarray(train_labels).astype('float32')\n","y_test = np.asarray(test_labels).astype('float32')"],"execution_count":0,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"sWEyG_LPZqm5","colab_type":"text"},"source":["A validation set will be created from the original training dataset."]},{"cell_type":"code","metadata":{"id":"eyMWZeDPZqFL","colab_type":"code","colab":{}},"source":["x_val=x_train[:10000]\n","partial_x_train=x_train[10000:]\n","y_val=y_train[:10000]\n","partial_y_train=y_train[10000:]"],"execution_count":0,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"2g1y_qa3Ww_t","colab_type":"text"},"source":["## Neural network 1"]},{"cell_type":"markdown","metadata":{"id":"n07znnyscOd4","colab_type":"text"},"source":["- two hidden layers\n","- number of units: 16\n","- loss function: binary_crossentropy\n","- activation function: relu"]},{"cell_type":"code","metadata":{"id":"x9n_9rISU6vO","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":255},"outputId":"bf8812cd-cf45-4028-dcde-b72bea972759","executionInfo":{"status":"ok","timestamp":1577083277762,"user_tz":-60,"elapsed":619,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["from keras import models \n","from keras import layers\n","\n","model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.summary()"],"execution_count":50,"outputs":[{"output_type":"stream","text":["Model: \"sequential_11\"\n","_________________________________________________________________\n","Layer (type)                 Output Shape              Param #   \n","=================================================================\n","dense_33 (Dense)             (None, 16)                160016    \n","_________________________________________________________________\n","dense_34 (Dense)             (None, 16)                272       \n","_________________________________________________________________\n","dense_35 (Dense)             (None, 1)                 17        \n","=================================================================\n","Total params: 160,305\n","Trainable params: 160,305\n","Non-trainable params: 0\n","_________________________________________________________________\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"XGJ4KADPXd61","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":714},"outputId":"6ba1f356-b30a-4ea0-930c-11bb47b33d6e","executionInfo":{"status":"ok","timestamp":1577083321250,"user_tz":-60,"elapsed":42901,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model.compile(optimizer='rmsprop',\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=20,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":51,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/20\n","15000/15000 [==============================] - 3s 189us/step - loss: 0.5178 - acc: 0.7827 - val_loss: 0.3897 - val_acc: 0.8677\n","Epoch 2/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.3117 - acc: 0.9014 - val_loss: 0.3132 - val_acc: 0.8842\n","Epoch 3/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.2273 - acc: 0.9267 - val_loss: 0.2783 - val_acc: 0.8904\n","Epoch 4/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.1802 - acc: 0.9419 - val_loss: 0.3184 - val_acc: 0.8713\n","Epoch 5/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.1473 - acc: 0.9527 - val_loss: 0.2894 - val_acc: 0.8849\n","Epoch 6/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.1235 - acc: 0.9612 - val_loss: 0.2987 - val_acc: 0.8838\n","Epoch 7/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.1020 - acc: 0.9688 - val_loss: 0.3177 - val_acc: 0.8818\n","Epoch 8/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.0860 - acc: 0.9753 - val_loss: 0.3386 - val_acc: 0.8796\n","Epoch 9/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0710 - acc: 0.9802 - val_loss: 0.3840 - val_acc: 0.8717\n","Epoch 10/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0603 - acc: 0.9836 - val_loss: 0.3818 - val_acc: 0.8791\n","Epoch 11/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0491 - acc: 0.9879 - val_loss: 0.3992 - val_acc: 0.8800\n","Epoch 12/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0421 - acc: 0.9897 - val_loss: 0.4292 - val_acc: 0.8769\n","Epoch 13/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0345 - acc: 0.9923 - val_loss: 0.4551 - val_acc: 0.8741\n","Epoch 14/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0290 - acc: 0.9933 - val_loss: 0.4798 - val_acc: 0.8743\n","Epoch 15/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0214 - acc: 0.9967 - val_loss: 0.5523 - val_acc: 0.8595\n","Epoch 16/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0190 - acc: 0.9971 - val_loss: 0.5464 - val_acc: 0.8710\n","Epoch 17/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0148 - acc: 0.9983 - val_loss: 0.5776 - val_acc: 0.8690\n","Epoch 18/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0147 - acc: 0.9967 - val_loss: 0.6090 - val_acc: 0.8683\n","Epoch 19/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0078 - acc: 0.9996 - val_loss: 0.6693 - val_acc: 0.8581\n","Epoch 20/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0084 - acc: 0.9991 - val_loss: 0.6764 - val_acc: 0.8676\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"lkBFLiUMaV2l","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":34},"outputId":"93ba6707-ece4-4bf7-f4bc-55e6034050eb","executionInfo":{"status":"ok","timestamp":1577083329992,"user_tz":-60,"elapsed":657,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["history_dict= history.history\n","history_dict.keys()"],"execution_count":52,"outputs":[{"output_type":"execute_result","data":{"text/plain":["dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])"]},"metadata":{"tags":[]},"execution_count":52}]},{"cell_type":"code","metadata":{"id":"-XJ-1Rvpa0Jl","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":280},"outputId":"6ebbb51e-abc2-4c03-f1dc-465581969487","executionInfo":{"status":"ok","timestamp":1577083331933,"user_tz":-60,"elapsed":1245,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["def visualize_history(history_dict):\n","  loss_values=history_dict['loss']\n","  val_loss_values = history_dict['val_loss']\n","  acc_values= history_dict['acc']\n","  val_acc_values= history_dict['val_acc']\n","\n","  epochs = range(1,len(loss_values)+1)\n","\n","  plt.figure(figsize=(13,4))\n","  plt.subplot(121)\n","  plt.plot(epochs, loss_values, 'bo', label='training loss')\n","  plt.plot(epochs, val_loss_values, 'b', label = 'validation loss')\n","  plt.xlabel('Epochs')\n","  plt.ylabel('Loss')\n","  plt.legend(frameon=False)\n","\n","  plt.subplot(122)\n","  plt.plot(epochs, acc_values, 'bo', label='training acc.')\n","  plt.plot(epochs, val_acc_values, 'b', label = 'validation acc.')\n","  plt.xlabel('Epochs')\n","  plt.ylabel('Accuracy')\n","  plt.legend(frameon=False)\n","\n","  plt.show()\n","\n","visualize_history(history_dict)"],"execution_count":53,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAwgAAAEHCAYAAAAKzGHrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nov8Hh7f9dYLVsFAAAAAASUVORK5CYII=\n","text/plain":["<Figure size 936x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"markdown","metadata":{"id":"wN_nWDhKdAsp","colab_type":"text"},"source":["You can clearly see overfitting. In the next cell, the same model will be trained again, but only for 4 epochs to avoid the overfitting."]},{"cell_type":"code","metadata":{"id":"Kgln9NlSb3N2","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":170},"outputId":"92418b46-062e-4d9e-eb52-acb5d2da0a43","executionInfo":{"status":"ok","timestamp":1577082693305,"user_tz":-60,"elapsed":9887,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer='rmsprop',\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=4,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":40,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/4\n","15000/15000 [==============================] - 3s 184us/step - loss: 0.5271 - acc: 0.7747 - val_loss: 0.4149 - val_acc: 0.8422\n","Epoch 2/4\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.3168 - acc: 0.8987 - val_loss: 0.3118 - val_acc: 0.8881\n","Epoch 3/4\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.2339 - acc: 0.9267 - val_loss: 0.2867 - val_acc: 0.8872\n","Epoch 4/4\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.1852 - acc: 0.9424 - val_loss: 0.2778 - val_acc: 0.8892\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"xZ1FtaGUa7p7","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":51},"outputId":"c4d1a495-2706-4e38-a107-c492c5667ebc","executionInfo":{"status":"ok","timestamp":1577082775395,"user_tz":-60,"elapsed":6436,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["results = model.evaluate(x_test,y_test)\n","results"],"execution_count":41,"outputs":[{"output_type":"stream","text":["25000/25000 [==============================] - 6s 238us/step\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["[0.29326927030563354, 0.881]"]},"metadata":{"tags":[]},"execution_count":41}]},{"cell_type":"markdown","metadata":{"id":"8GHwx5zxeAdx","colab_type":"text"},"source":["## Neural network 2\n","\n","Same as NN1\n","- two hidden layers\n","- number of units: 16\n","- loss function: binary_crossentropy\n","- activation function: relu\n","\n","Only the optimizer is different."]},{"cell_type":"code","metadata":{"id":"LN935sPOdhYL","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":714},"outputId":"a793f55d-b7ff-416f-dd46-be821c6be6f8","executionInfo":{"status":"ok","timestamp":1577083390715,"user_tz":-60,"elapsed":43169,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["from keras import optimizers\n","\n","model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer=optimizers.RMSprop(lr=0.001),\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=20,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":54,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/20\n","15000/15000 [==============================] - 3s 183us/step - loss: 0.5497 - acc: 0.7806 - val_loss: 0.4295 - val_acc: 0.8674\n","Epoch 2/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.3484 - acc: 0.8981 - val_loss: 0.3314 - val_acc: 0.8819\n","Epoch 3/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.2516 - acc: 0.9231 - val_loss: 0.3238 - val_acc: 0.8706\n","Epoch 4/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.1997 - acc: 0.9381 - val_loss: 0.2792 - val_acc: 0.8883\n","Epoch 5/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.1617 - acc: 0.9500 - val_loss: 0.2767 - val_acc: 0.8883\n","Epoch 6/20\n","15000/15000 [==============================] - 2s 136us/step - loss: 0.1344 - acc: 0.9586 - val_loss: 0.2954 - val_acc: 0.8846\n","Epoch 7/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.1124 - acc: 0.9672 - val_loss: 0.2959 - val_acc: 0.8849\n","Epoch 8/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0943 - acc: 0.9739 - val_loss: 0.3283 - val_acc: 0.8800\n","Epoch 9/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0803 - acc: 0.9786 - val_loss: 0.3676 - val_acc: 0.8759\n","Epoch 10/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0655 - acc: 0.9846 - val_loss: 0.4032 - val_acc: 0.8709\n","Epoch 11/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0548 - acc: 0.9870 - val_loss: 0.3734 - val_acc: 0.8785\n","Epoch 12/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0453 - acc: 0.9897 - val_loss: 0.4639 - val_acc: 0.8596\n","Epoch 13/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.0361 - acc: 0.9928 - val_loss: 0.4588 - val_acc: 0.8678\n","Epoch 14/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.0303 - acc: 0.9943 - val_loss: 0.4536 - val_acc: 0.8743\n","Epoch 15/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0240 - acc: 0.9955 - val_loss: 0.4846 - val_acc: 0.8717\n","Epoch 16/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0195 - acc: 0.9965 - val_loss: 0.5113 - val_acc: 0.8704\n","Epoch 17/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0128 - acc: 0.9985 - val_loss: 0.5469 - val_acc: 0.8678\n","Epoch 18/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0116 - acc: 0.9986 - val_loss: 0.5835 - val_acc: 0.8680\n","Epoch 19/20\n","15000/15000 [==============================] - 2s 136us/step - loss: 0.0096 - acc: 0.9987 - val_loss: 0.6178 - val_acc: 0.8649\n","Epoch 20/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0051 - acc: 0.9998 - val_loss: 0.6426 - val_acc: 0.8650\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"pCX6OIRVdhbs","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":279},"outputId":"01e70880-14a9-469a-d987-e029efa89f98","executionInfo":{"status":"ok","timestamp":1577083420198,"user_tz":-60,"elapsed":1105,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["history_dict= history.history\n","visualize_history(history_dict)"],"execution_count":56,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAwgAAAEGCAYAAADBkLJOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOzdeXhU5dnH8e9DWGxYgyICQoKKCoGw\nJKIVcSmiCBYKioDYioooxe1VqSgqiuJeRNyjdQ9FwL0FFxQFWlGCLIJFUQi7GGQ3bCH3+8czCSEk\nIWRmMknm97muc82cM2e5E8Kccz+rMzNEREREREQAqkQ6ABERERERKT+UIIiIiIiISB4lCCIiIiIi\nkkcJgoiIiIiI5FGCICIiIiIieapGOoDDddRRR1lCQkKkwxARKbfmzZu30cwaRDqOSNP9QkSkaMXd\nKypcgpCQkEB6enqkwxARKbeccysjHUN5oPuFiEjRirtXqImRiIiIiIjkUYIgIiIiIiJ5lCCIiIiI\niEgeJQgiIiIiIpJHCYKIiIiIiORRgiAiImXCOfeSc+4X59ziIj53zrnxzrkfnXOLnHMd8n12uXNu\nWWC5vOyiFhEpf9LSICEBqlTxr2lpoT2/EgQRESkrrwDdivn8AqBFYBkCPAvgnKsPjAJOBToCo5xz\ncWGNVEQqvWAfsiN1fFoaDBkCK1eCmX8dMiS0SYISBBGRciQrCx5/HD77LNKRhJ6ZzQQ2FbNLL+A1\n8+YA9ZxzjYDzgU/MbJOZbQY+ofhEQ0SkWME+ZEfy+JEj/b0iv6wsvz1UlCCEwJYtW3jmmWdKdWz3\n7t3ZsmVLsfvcfffdTJ8+vVTnLyghIYGNGzeG5FwiEjrbt8NDD/lSpJtvhg8+iHREEdEEWJ1vfU1g\nW1HbD+KcG+KcS3fOpWdmZoYtUBGJvGBK8IN9yI7k8atWHd720ojKBCHU7baKSxCys7OLPXbq1KnU\nq1ev2H1Gjx7NueeeW+r4RKT82rwZRo+G+Hi4/XZITobZs30tghw+M0s1sxQzS2nQoEGkwxGRQ4hU\nM5tgH7IjeXyzZoe3vTSiLkEIR7utESNG8NNPP9GuXTuGDx/O559/TufOnenZsyetWrUC4E9/+hPJ\nyckkJiaSmpqad2xuiX5GRgYtW7bk6quvJjExkfPOO4+dO3cCMGjQIKZMmZK3/6hRo+jQoQNt2rRh\n6dKlAGRmZtK1a1cSExMZPHgw8fHxh6wpGDt2LK1bt6Z169aMGzcOgN9++40ePXrQtm1bWrduzZtv\nvpn3M7Zq1YqkpCRuvfXW0v+yRASAjRt9SVFCAowaBZ07w9dfw7Rp0KlTpKOLmLVA03zrxwa2FbVd\nRCIsmELXSDazCfYhO5LHjxkDsbEHbouN9dtDxswq1JKcnGzBiI8383+GBy7x8aU/54oVKywxMTFv\nfcaMGRYbG2vLly/P2/brr7+amVlWVpYlJibaxo0bA/HEW2Zmpq1YscJiYmJs/vz5ZmbWt29fe/31\n183M7PLLL7fJkyfn7T9+/HgzM3v66aftqquuMjOzYcOG2QMPPGBmZtOmTTPAMjMzC/n5/fXS09Ot\ndevWtmPHDtu+fbu1atXKvvnmG5syZYoNHjw4b/8tW7bYxo0b7cQTT7ScnBwzM9u8eXPpf1kiUW79\nerNbbjGLjTVzzqxvX7MFC0J7DSDdysH3dWELkAAsLuKzHsA0wAGnAV8HttcHVgBxgWUFUP9Q1wr2\nfiEixXvjDf9dlv95KjbWby+JYJ7JnCv8WOfKJvbycHx8vP954+NLflx+xd0roq4GoSzabQF07NiR\n5s2b562PHz+etm3bctppp7F69WqWLVt20DHNmzenXbt2ACQnJ5ORkVHoufv06XPQPrNnz6Z///4A\ndOvWjbi44gf4mD17Nr1796ZmzZrUqlWLPn36MGvWLNq0acMnn3zCbbfdxqxZs6hbty5169bliCOO\n4KqrruLtt98mtmDaKiKHtHo13HADNG/umw/16QOLF8OkSdC2baSjKxvOuX8CXwInOefWOOeucs5d\n65y7NrDLVGA58CPwAvBXADPbBNwHzA0sowPbRCRIkWzHH8lmNgMHQmqqb97pnH9NTfXbK8rxGRmQ\nk+NfS3pcSVUN7enKv2bNfBVWYdtDqWbNmnnvP//8c6ZPn86XX35JbGwsZ599Nrt27TromBo1auS9\nj4mJyWtiVNR+MTExh+zjcLhOPPFEvvnmG6ZOncqdd95Jly5duPvuu/n666/59NNPmTJlCk899RSf\nVcYhVkTCYMUK3/n45Zd9GdFf/uL7GpxwQqQjK3tmNuAQnxswrIjPXgJeCkdcItEqt4lP7kN+bhMf\nKNkDZ7CFrsE8k40Zc2DscPjNbAYODO7BOtLHh1PU1SCEo91W7dq12b59e5Gfb926lbi4OGJjY1m6\ndClz5swp/cWK0KlTJyZNmgTAxx9/zObNm4vdv3Pnzrz77rtkZWXx22+/8c4779C5c2fWrVtHbGws\nl112GcOHD+ebb75hx44dbN26le7du/P444+zcOHCkMcvUtn88AMMGgQtWsArr8DgwbBsGfzjH9GZ\nHIhIeESyBiDYUvxgnsmCLYGX4kVdDULuH87IkT7DbdbM/yEG8wd15JFH0qlTJ1q3bs0FF1xAjx49\nDvi8W7duPPfcc7Rs2ZKTTjqJ0047LYifoHCjRo1iwIABvP766/z+97/nmGOOoXbt2kXu36FDBwYN\nGkTHjh0BGDx4MO3bt+ejjz5i+PDhVKlShWrVqvHss8+yfft2evXqxa5duzAzxo4dG/L4RSqLrVvh\nllt8jUGNGnD99XDrrdCk0EE5RURKL9I1AMGW4gf7TFaeS+ArOudrdCuOlJQUS09Pj3QY5c7u3buJ\niYmhatWqfPnllwwdOpQFCxZEOiyRqPLxx3DVVbBuHdx4I9x2GzRsWPZxOOfmmVlK2V+5fNH9Qiq7\nhITCm+jEx/t26eE+HnySEspCVyk7xd0roq6JUWW1atUqTjnlFNq2bcsNN9zACy+8EOmQRKLG9u1w\nzTVw/vlQuzbMmQNjx0YmORCRiiWYJkKhqAEIttl1uDvLSmSEtYmRc64b8AQQA7xoZg8Vss8lwD2A\nAQvN7NJwxlRZtWjRgvnz50c6DJGo89lncOWV/oY8fLif9OyIIyIdlYhUBME2EQp24JVwNLuWyiFs\nNQjOuRjgaeACoBUwwDnXqsA+LYDbgU5mlgjcFK54RERCaccOuO466NIFqlf3sx8/8oiSA5FoVNpa\ngGA7CasGQMIlnE2MOgI/mtlyM9sDTAR6FdjnauBpM9sMYGa/hDEeEZGQmDXLz13wzDNw002wYAGc\nfnqkoxKRSAhmNuBgmwhpJB8Jl3AmCE2A1fnW1wS25XcicKJz7j/OuTmBJkkHcc4Ncc6lO+fSMzMz\nwxSuiEjxsrLg//4PzjrLr3/+uZ/0THMHikSvYGoBgh0mFFQDIOER6U7KVYEWwNnAAOAF51y9gjuZ\nWaqZpZhZSoMGDco4RBER+O9/oV07GDcO/vpXWLQIzjwz0lGJSKQFUwsQjrmZREIhnAnCWqBpvvVj\nA9vyWwO8b2Z7zWwF8AM+Yaj0atWqBcC6deu4+OKLC93n7LPP5lBD9I0bN46sfEUX3bt3Z8uWLUHH\nd8899/DYY48FfR6Rim7XLvjb36BzZ9izBz79FJ56CvJNli4iUSyYWgA1EZLyKpwJwlyghXOuuXOu\nOtAfeL/APu/iaw9wzh2Fb3K0PIwxlTuNGzdmypQppT6+YIIwdepU6tU7qBJGREph7lzo0AEefdTP\nhLxoEfzhD5GOSkRCLZihRoOtBVATISmPwpYgmFk2cB3wEfA/YJKZLXHOjXbO9Qzs9hHwq3PuO2AG\nMNzMfg1XTOEyYsQInn766bz13NL3HTt20KVLFzp06ECbNm147733Djo2IyOD1q1bA7Bz50769+9P\ny5Yt6d27Nzt37szbb+jQoaSkpJCYmMioUaMAGD9+POvWreOcc87hnHPOASAhIYGNGzcCMHbsWFq3\nbk3r1q0ZN25c3vVatmzJ1VdfTWJiIuedd94B1ynMggULOO2000hKSqJ3795s3rw57/qtWrUiKSmJ\n/v37A/DFF1/Qrl072rVrR/v27dm+fXupfqcikbR7t28//Pvf+zkOPvoInn8e6tSJdGQiEmrBdDIG\n1QJI5VTpZlLOHVEklHLbHRdl/vz53HTTTXzxxRcAtGrVio8++ohGjRqRlZVFnTp12LhxI6eddhrL\nli3DOUetWrXYsWMHGRkZXHjhhSxevJixY8eyePFiXnrpJRYtWkSHDh2YM2cOKSkpbNq0ifr167Nv\n3z66dOnC+PHjSUpKIiEhgfT0dI466iiAvPWVK1cyaNAg5syZg5lx6qmn8sYbbxAXF8cJJ5xAeno6\n7dq145JLLqFnz55cdtllB/xM99xzD7Vq1eLWW28lKSmJJ598krPOOou7776bbdu2MW7cOBo3bsyK\nFSuoUaMGW7ZsoV69evzxj39kxIgRdOrUiR07dnDEEUdQtWpYp9sQCamlS2HAAP89csUVfsKzilYp\np5mUPc2kLCURitmERSoizaQcZu3bt+eXX35h3bp1LFy4kLi4OJo2bYqZcccdd5CUlMS5557L2rVr\n2bBhQ5HnmTlzZt6DelJSEklJSXmfTZo0iQ4dOtC+fXuWLFnCd999V2xMs2fPpnfv3tSsWZNatWrR\np08fZs2aBUDz5s1p164dAMnJyWQU8w24detWtmzZwlmBYVsuv/xyZs6cmRfjwIEDeeONN/KSgE6d\nOnHzzTczfvx4tmzZouRAKgwzePFFSE6G1avh3XfhpZcqXnIgIocn2KFGRSqjSvf0VlxJfzj17duX\nKVOm8PPPP9OvXz8A0tLSyMzMZN68eVSrVo2EhAR27dp12OdesWIFjz32GHPnziUuLo5BgwaV6jy5\natSokfc+JibmkE2MivLvf/+bmTNn8sEHHzBmzBi+/fZbRowYQY8ePZg6dSqdOnXio48+4uSTTy51\nrCJlYdMm36Tgrbf8xGevvQaNG0c6KhEpC8HORixSGakGIUT69evHxIkTmTJlCn379gV86fvRRx9N\ntWrVmDFjBisL+wbK58wzz2TChAkALF68mEWLFgGwbds2atasSd26ddmwYQPTpk3LO6Z27dqFtvPv\n3Lkz7777LllZWfz222+88847dO7c+bB/rrp16xIXF5dX+/D6669z1llnkZOTw+rVqznnnHN4+OGH\n2bp1Kzt27OCnn36iTZs23HbbbZxyyiksXbr0sK8pUpa++MJPevbee34m5I8/VnIgUtFEspOxSGVU\n6WoQIiUxMZHt27fTpEkTGjVqBMDAgQP54x//SJs2bUhJSTlkSfrQoUO54ooraNmyJS1btiQ5ORmA\ntm3b0r59e04++WSaNm1Kp06d8o4ZMmQI3bp1o3HjxsyYMSNve4cOHRg0aBAdO3YEYPDgwbRv377Y\n5kRFefXVV7n22mvJysriuOOO4+WXX2bfvn1cdtllbN26FTPjhhtuoF69etx1113MmDGDKlWqkJiY\nyAUXXHDY1xMpC3v3wr33wgMPwAknwJdfQkrUt9oXqXhyOxnnDuiX28kYStZROHefkSN9s6JmzXxy\noE7GEs0qXSdlEZFDWb4cLr0UvvoKrrwSnngCAlOTVArqpOzpfhEd1MlYpHSKu1eoBkFEosobb/iZ\nkGNiYNIkCLQIFJEKSp2MRUJPfRBEJCps2waXXQZ//rMfunjhQiUHIpVBMDMZi0jhoiJBCKbzkohU\nfHPm+KRg4kQYPRpmzNDDg0hloU7GIqFX6ROEYGdIFJGKa98+uP9+OOMM//9/5ky46y7fvEhEyo9g\nCvI0k7FI6FX6TsrqvCQSnVav9k2KZs70MyM/+yzUrRvpqMqGOil76qRcMRQchQh8DYAe8kXCK6pn\nUlbnJZHoM2kSJCXBN9/4Sc/S0qInORCpaEaOPDA5AL8+cmRk4hGRKEgQ1HlJJHps3eo7IffrByed\nBAsW+HXnIh2ZiBRFBXki5U+lTxDUeUkkOsyc6WsN/vlPuOcemD0bjj8+0lGJyKGoIE+k/Kn0CYI6\nL4lUbrt3w223wdlnQ/Xq8J//wKhRUFWzvIhUCCrIEyl/ouIWOnCgEgKRymjJEt8RecEC38nx73+v\nXDMii0SD3PvzyJG+WVGzZj450H1bJHIqfQ2CiFQ+OTkwfjwkJ8PatfDee/D880oORCIp2KFKMzL8\n/+2MDCUHIpEWFTUIIlJ5rF0LV1wBn3wCF14IL74IDRtGOiqR6FZwqNLcOYdAD/siFZFqEESkwpgy\nBdq08f0MnnsO3n9fyYFIeaChSkUqFyUIIlLubdsGl18OffvCCSfA/PlwzTUavlSkvNBQpSKVixIE\nESnXZs2Ctm3hjTfg7rt97cGJJ0Y6KhHJT0OVilQuShBEpFzaswduvx3OOgtiYvy8BvfeC9WqRToy\nESlIQ5WKVC5KEESkXDGD6dOhY0d46CG46io/jOnvfx/pyESkKJpzSKRyCWuC4Jzr5pz73jn3o3Nu\nRCGfD3LOZTrnFgSWweGMR0TKtzlzoEsX6NoVNm+Gd96BF17Q8KUiFYGGKhWpPMKWIDjnYoCngQuA\nVsAA51yrQnZ908zaBZYXwxWPiJRf334LvXr5WoIlS+CJJ+CHH+BPf4p0ZBJKJSg0infOfeqcW+Sc\n+9w5d2y+z/blK0x6v2wjjw7BzGMgIpVLOOdB6Aj8aGbLAZxzE4FewHdhvKaIVCA//QSjRsGECVCn\nDtx/P9x4o2oMKqN8hUZdgTXAXOfc+2aW/57wGPCamb3qnPsD8CDw58BnO82sXZkGHUU0j4GI5BfO\nJkZNgNX51tcEthV0UaC0aIpzrmlhJ3LODXHOpTvn0jMzM8MRq4iUobVr4dpr4eST4e234W9/g+XL\n/ZjpSg4qrbxCIzPbA+QWGuXXCvgs8H5GIZ9LmGgeAxHJL9KdlD8AEswsCfgEeLWwncws1cxSzCyl\nQYMGZRqgiITOr7/C8OF+LoN//MOXUP70k++MXL9+pKOTMCtJodFCoE/gfW+gtnPuyMD6EYGCojnO\nuSIbn6lAqXQ0j4GI5BfOBGEtkL9G4NjAtjxm9quZ7Q6svggkhzEeEYmQ7dth9Gho3hz+/ne45BL4\n/nt4+mlo1CjS0Uk5citwlnNuPnAW/p6xL/BZvJmlAJcC45xzxxd2AhUolY7mMRCR/MKZIMwFWjjn\nmjvnqgP9gQM6ljnn8j8a9AT+F8Z4RKSM7doFjz8Oxx3n+xqce67vkPzqq36bRJWSFBqtM7M+ZtYe\nGBnYtiXwujbwuhz4HGhfBjFHDc1jICL5hS1BMLNs4DrgI/yD/yQzW+KcG+2c6xnY7Qbn3BLn3ELg\nBmBQuOIRkbKTkwMvvwwtWsDNN0O7dvD1176/QWJipKOTCClJodFRzrnc+9LtwEuB7XHOuRq5+wCd\n0IAXIaV5DEQkv3COYoSZTQWmFth2d773t+NvAiJSSSxYAH/9K3z5JZx6qq8t+MMfIh2VRJqZZTvn\ncguNYoCXcguNgHQzex84G3jQOWfATGBY4PCWwPPOuRx8wdZDBUY/khAYOFAJgYh4YU0QRCR6bNsG\nd98NTz4JRx4Jr7wCf/mLL40UgRIVGk0BphRy3H+BNmEPUEREACUIIhIkM3jzTd+U6Oef/fClY8ZA\nXFykIxMREZHSUIIgIqW2dCkMGwaffQbJyfDee3DKKZGOSkRERIIR6XkQRKQCyp1AKSkJ5s3zw5V+\n9ZWSA5FISkuDhASoUsW/pqVFOiIRqahUgyBSjm3bBjVq+KW8eP99uOEGWLnS9zF45BFo2DDSUYlE\nt7Q0P/Fg7mzIK1f6dVDHYxE5fEoQRMqpBQugY0fYu9d3+m3UCBo33v+a/32jRn4JZyKRkeETgw8+\ngFat4PPP4ayzwnc9ESm5kSP3Jwe5cmv6lCCIyOFSgiBSDpnBLbdAnTr+oXz9er+sWwfffeff79t3\n8HH16x+cRBRcjjkGqlcveSy7d/vZj++/349I9MgjcNNNUK1a6H5eEQnOqlWHt11EpDhKEETKoX/9\ny3f8feop3wm4oJwc2LjRJwy5iUPB16VL/fvs7IOPP/roohOIJk38a4MGMGOGv/4PP8BFF/lZkZs2\nPfh8IhJZzZr5ZkWFbRcROVxKEETKmb174dZb4eST97chLqhKFf+Qf/TRfpbiouRPJNatg7Vr97/P\nXb75BjZs8LUW+cXE+FqK44+HqVPhggtC9zOKSGiNGXNgHwSA2Fi/XUTkcClBEClnnnvOl9h/8EHw\nzXhKmkjs3euThILJQ/36MHQo/O53wcUhIuGV289g5EjfrKhZM58cqP+BiJSGEgSRcmTzZrjnHujS\nBXr0KLvrVqsGxx7rFxGpmAYOVEIgIqGheRBEypH77/dJwt//7jsEi4iIiJQ1JQgi5cSPP8KTT8KV\nV0LbtpGORkRERKKVEgSRcmLECD/86H33RToSERERiWZKEETKgVmz4K23fJLQqFGkoxEREZFoFjUJ\nwmefwWuvRToKkYPl5MDNN/sOwjffHOloREREJNpFTYLw5JNw/fWwaVOkIxE50IQJkJ4ODzzgxy0X\nkeiUlgYJCX544oQEvy4iEglRkyDcdx9s3w6PPhrpSET2y8qC22+H5GQNTygSzdLS/ERnK1f6SQtX\nrvTrShJEJBKiJkFo3RoGDIDx4+HnnyMdjYg3diysWeNfq0TN/0YRKWjkyANnQQa/PnJkZOIRkegW\nVY8k99wDu3fDgw9GOhIRWL8eHnoI+vSBM8+MdDQiEkmrVh3edhGRcIqqBKFFCxg0CJ57Tl+6Enl3\n3w179sDDD0c6EhGJtGbNDm+7iEg4RVWCAP6hDPyMtSKRsnAh/OMfvuP8CSdEOhoRibQxYw4epCA2\n1m8XESlrYU0QnHPdnHPfO+d+dM6NKGa/i5xz5pxLCWc84EtjrrkGXnrJz1wrUtbM4JZbIC4O7rwz\n0tGISHkwcCCkpkJ8PDjnX1NTNXiBiERG2BIE51wM8DRwAdAKGOCca1XIfrWBG4GvwhVLQXfc4Wes\nveeesrqiVFQbNvjRREJp6lT49FP/9xcXF9pzi0jFNXAgZGT4uVEyMpQciEjkhLMGoSPwo5ktN7M9\nwESgVyH73Qc8DOwKYywHOOYY37RjwgRYvLisrioViZkfXvCkk3wToOHDYdu24M+7dy/ceiuceCJc\ne23w5xMREREJtXAmCE2A1fnW1wS25XHOdQCamtm/izuRc26Icy7dOZeemZkZkuD+9jeoXXt/nwSR\nXJs2+SFxL7sMEhP962OP+Yf6V1/1pXullZoKS5f6+TiqVQtdzCIiIiKhErFOys65KsBY4JZD7Wtm\nqWaWYmYpDRo0CMn1jzwSbr4Z3nkH5s0LySmlEpg+HZKS4K23fOfAmTPh5Zfhq6/8zKaDBsHpp8Pc\nuYd/7i1bfLOic86BP/4xxIGLiIiIhEg4E4S1QNN868cGtuWqDbQGPnfOZQCnAe+XRUflXP/3f1C/\nvjqKCuzcCTfdBF27+pqlOXN8X5WYGP95x47w3//6GoSMDL9+1VW+j0JJPfAA/Por/P3vvhOiiIiI\nSHkUzgRhLtDCOdfcOVcd6A+8n/uhmW01s6PMLMHMEoA5QE8zSw9jTAeoUwduuw0+/BBmzy6rq0p5\nM38+pKTAE0/4vinz5kFy8sH7VakCf/kL/PCD75Pw+uu+2dHjj/u+BcVZvtyff9AgaN8+LD+GiIiI\nSEiELUEws2zgOuAj4H/AJDNb4pwb7ZzrGa7rHq7rrvOdlkeO9B1TJXrs2+dnMj71VNi82SeK48cf\nPBZ5QXXqwCOPwLff+uZGN98MbdvCJ58UfcyIEVC1qubfkMrBOXe9c05jcImIVFJh7YNgZlPN7EQz\nO97MxgS23W1m7xey79llWXuQKzbWJwczZ/r25xIdVqyAs8+G22+HXr38w/755x/eOU46yQ9Z+sEH\nfkbk886D3r19bUF+//kPTJ7sa6saNw7ZjyASSQ2Buc65SYH5btRoTkSkEom6mZQLc/XVfgI11SJU\nfmbwyiu+xH/RInjtNZg0yXdaLw3n4MILYckSePBBX4vQqhXcdRf89psf8ejmm31icMshu+OLVAxm\ndifQAvgHMAhY5px7wDl3fEQDExGRkFCCANSoAaNG+ZFp3j+obkMqi40b4eKL4YorfD+ARYvgz38O\nTYfhGjV8M6Lvv/fXuP9+OPlk36fh6699B+WaNYO/jkh5YWYG/BxYsoE4YIpz7pGIBiYiIkFTghDw\nl79Aixa+5DeYce6lfJo2Ddq08c2BHn4YPvsM4uNDf50mTeCNN3yn9wYN4JlnoEMHn4iIVBbOuRud\nc/OAR4D/AG3MbCiQDFwU0eBERCRoShACqlaFe+/1bdEnTYp0NBIqWVkwbBh07+6bEc2d6yfJyx2+\nNFw6dfLXmjzZ/z1V0f80qVzqA33M7Hwzm2xmewHMLAe4MLKhiYhIsPTYkk+/fr6UedQoyM6OdDQS\nrGXL4LTTfCn+//0fpKf7vgdlJSbGNzc6Xq2ypfKZBmzKXXHO1XHOnQpgZv+LWFQiIhISShDyqVIF\n7rvPj3P/2muRjkaC8c47fm6DtWt986KxY+GIIyIdlUil8SywI9/6jsA2ERGpBJQgFNCzJ5xyim9u\ntHt3pKORw5Wd7YcT7dPHD0P6zTfQrVukoxKpdFygkzKQ17So6iEP8kOifu+c+9E5N6KQz+Odc586\n5xY55z53zh2b77PLnXPLAsvlIftJRETkIEoQCnDOj0CzahW8+KLflpYGCQm+hiEhwa9L+fPzz3Du\nuX4Ss6FDYdas8HREFhGWO+ducM5VCyw3AsuLO8A5FwM8DVwAtAIGOOdaFdjtMeA1M0sCRgMPBo6t\nD4wCTgU6AqM0UZuISPgoQShE165w5pk+UXjpJRgyBFau9GPor1zp15UklC+zZ/vRgr7+2jcPe+YZ\nP/SoiITFtcDpwFpgDf7BfcghjukI/Ghmy81sDzAR6FVgn1bAZ4H3M/J9fj7wiZltMrPNwCdAuawb\nVIGSiFQGShAK4RyMGeNLpG+5xY+Ek19Wlp9UTSLPDB5/3M+KXLMmfPWVhhQVCTcz+8XM+pvZ0WbW\n0MwuNbNfDnFYE2B1vvU1gcJXyKsAACAASURBVG35LQT6BN73Bmo7544s4bEAOOeGOOfSnXPpmZmZ\nJf2RQiItTQVKIlI5lChBcM4d75yrEXh/dqBquV54Q4usM87wbde3bCn881WryjYeOdi2bXDJJX6m\n4p49/ShFbdpEOiqRys85d4Rzbphz7hnn3Eu5SwhOfStwlnNuPnAWvoZi3+GcwMxSzSzFzFIaNGgQ\ngpBKbuRIFSiJSOVQ0hqEt4B9zrkTgFSgKTAhbFGVE/fdV/RnzZqVXRxysCVLoGNHP1rRo4/CW29B\n3bqRjkokarwOHINv+vMFcCyw/RDHrMXfO3IdG9iWx8zWmVkfM2sPjAxs21KSY8uDogqOVKAkIhVN\nSROEHDPLxlf5Pmlmw4FG4QurfEhJ8UtBsbG+CZJExj//6ZODLVvg00/h1lt9szARKTMnmNldwG9m\n9irQA98PoThzgRbOuebOuepAf+D9/Ds4545yzuXel24HcmslPgLOc87FBTonnxfYVq4UVXCkAiUR\nqWhKmiDsdc4NAC4H/hXYVi08IZUvr7ziHz7r1PGv8fGQmgoDB0Y6suizZw9cfz1ceqnvkDx/Ppx1\nVqSjEolKewOvW5xzrYG6wNHFHRAoZLoO/2D/P2CSmS1xzo12zvUM7HY28L1z7gegITAmcOwm4D58\nkjEXGB3YVq6MGeMLkPJTgZKIVESHHLc64Ar8qBVjzGyFc645voq50ktM9A+kb78N69bBMcdEOqLo\ntHq1728wZ47vOP7gg1AtKlJUkXIpNVCSfye+FqAWcNehDjKzqcDUAtvuzvd+CjCliGNfYn+NQrmU\nW3A0cqRvVtSsmU8OVKAkIhVNiRIEM/sOuAEgcFOobWYPhzOw8uSee2DiRDjxRD/5Vv7XE0+EFi2g\ndu1IR1l5TZ8OAwb4iesmT4aLL450RCLRK9AEaFtguNGZwHERDqlcGThQCYGIVHwlShCcc58DPQP7\nzwN+cc79x8xuDmNs5cYJJ8C//gX//jf88AP85z++Hfz+eUShUaMDk4bc5bjjVNJdWjt2wF13wRNP\nQKtWviPySSdFOiqR6GZmOc65vwGTIh2LiIiER0mbGNU1s23OucH4WS5HOecWhTOw8qZbN7/k2rkT\nfvrJJwy5y/ff+6ZIGzfu3y8mxicJLVvCqFG+7bwc2rRpfjbklSv966OP+nkORKRcmO6cuxV4E/gt\nd2N57BcgIiKHr6QJQlXnXCPgEgJDz0W73/0OWrf2S0GbNsGyZQcmDjNn+tmZJ02C7t3LPt6K4pdf\n4KabfA1Ny5Z+huROnSIdlYgU0C/wOizfNkPNjUREKoWSJgij8SNP/MfM5jrnjgOWhS+siq1+fTj1\nVL/kWr8eevTwE3o984yfXVP2M4NXX/UdkLdv9/0+RoyAGjUiHZmIFGRmzSMdgxxs71644go/P0ww\n4uP9UNK5S1ISVK8emhhFpGIoaSflycDkfOvLgYvCFVRl1KiRr0W45BK45hrfdOb++zV+P/imWtdc\n4+c06NQJXnjB1x6ISPnknPtLYdvN7LWyjkW8vXuhf3/fzPWKK+DII0t3npwcX/M9bZovtAFfUNO+\n/YFJwwkn6P4lUpmVtJPyscCTQG5jj1nAjWa2JlyBVUa1asH77/s29Q884JOEl16K3pKZvXth7Fhf\nW1C9Ojz7rK9ZqVLS2TlEJFJOyff+CKAL8A2gBCECsrP9yElvvw3jxsGNNwZ/TjM/VOvXX+9fXnwR\nxo/3n8fFwSmn+Jryjh39+4YNg7+uiJQPJW1i9DIwAegbWL8ssK1rcQc557oBTwAxwItm9lCBz6/F\nt2HdB+wAhgSGVK20qlb1E601b+7Hyl63zn+p16sX6cjKVno6DB4MCxdC797w5JPQpEmkoxKRkjCz\n6/OvO+fqARMjFE5Uy86Gyy7zQ0D//e+hSQ5g/8Sg8fHQt+/+a3333YFJw5gxvtYB/L7JyX5yuL17\n/f7Z2fvfl2Rb9eq+v163bnDuuT4REZGy5yz/WJ1F7eTcAjNrd6htBT6PAX7AJxFr8LNfDsifADjn\n6pjZtsD7nsBfzaxbYefLlZKSYunp6YeMuSJ44w248ko/HOq0adC0aaQjCr/ffts/dGnDhvDUU9Cn\nT6SjEqlcnHPzzCylDK9XDVhsZuVqIOLKdL8ozL598Je/wIQJ8MgjMHx42cfw229+Vvuvv4avvoIF\nC/zDfrVqvkCsatX970uybetWmDEDtmzxtcmnnrp/FMHkZD8yoIiERnH3ipLWIPzqnLsM+GdgfQDw\n6yGO6Qj8GOivgHNuItALyEsQcpODgJr4UTCixmWXQePGvgT9tNP8PAvtiky5Kr4PP4Rrr/VNq665\nBh56KPpqTkQqA+fcB+z/vq4CtELzIuRZu9Z/t4ezjf6+fTBokE8OHnwwMskB+OGnzzjDL6GSne0T\njg8/hI8+8s1QR43y/Sq6dvXJwnnn+b59IhIeJW3tfSV+iNOfgfXAxcCgQxzTBFidb31NYNsBnHPD\nnHM/AY8QmK25kH2GOOfSnXPpmZmZJQy5YvjDH/zEa1WqQOfO/suwssnM9MnQBRf44WFnzoTnnlNy\nIFKBPQb8PbA8CJxpZiMiG1LkZWfD9dfDscfCn/7kE4Vw2LfP1z6/8YZv4jOikv3mq1aF00+H0aN9\nrcQvv/hE6MILfe3CoEE+AWvXzv/sn38Oe/ZEOmqRyqVETYwKPdC5m8xsXDGfXwx0M7PBgfU/A6ea\n2XVF7H8pcL6ZXV7cdStrlfHatX4Y1MWLfR+FK6+MdER+uNG33/Y3oRUr/E0pJ8cvue+Les3/3sxX\nHd9+O9xxh4YuFQm3cDcxcs41B9ab2a7A+u+AhmaWEa5rlkZZ3i82b/aj1E2fDr16wccf+++9Rx6B\nq68O3eALOTlw1VXwyiv+Afquu0Jz3ooiJwcWLdpfuzB7tk/MatXyBW5du/rCttat1RxJ5FBC0cSo\nMDcDRSYIwFogf6v6YwPbijIReDaIeCq0Jk18yXrfvv7Lf+VKX61a1sPI7dvnhxt97TU/lnZWFhx/\nvG8HGhPjb3KH+1qtGlx0EbRqVbY/i4iEzWTg9Hzr+wLbTil898pt6VI/x01Ghh+Z7oor/PDNV1/t\nm1X+859++OYWLYK7Tk6OH+ntlVd8k5toSw7A31Patdtfe7B9O3z2mU8Wpk3zIwUC1KnjayFymz91\n7OhrsMuz7duhdu1IRyHiBZMgHOrRdS7QIlDStBboD1x6wAmca2FmuROu9SDKJ1+rUwf+9S/fPn/0\naD/EXGqqf8AOt8WLfVKQluZHVqpXD/78Z98B7ve/13jXInKAqmaW16jDzPY456JywOaPP/Y1B9Wr\n++YvuTO/H3+8L2x56SU/AWRSEtx7L9x8s29Cc7hycvwQ2f/4B9x5p08QxD9Q9+rll9yhWWfN8jUL\ns2f73xX4+2hy8v6EoVMnOOqoyMYOvvbjvff8wB2zZ/uJVK+9NtJRiQBmVqoFWFWCfbrjRzL6CRgZ\n2DYa6Bl4/wSwBFgAzAASD3XO5ORkq+xycszuvdcMzLp2Ndu6NTzX+flns7Fjzdq399eqWtWsZ0+z\nKVPMdu4MzzVFJPyAdCvld3tJFuCT3O/xwHov4NNwXrM0SzjvFzk5Zk88YValillSkllGRtH7rl1r\n1ru3/57t0MFs/vzDv9bQof7422/361Iyv/5q9sEHZrfdZtapk1n16v73CGYnn2w2eLDZK6+Y/fRT\n2f5eN20ye/RRs2bNfCwJCWannWbmnNnEiWUXh0S34u4Vh7oJbAe2FbJsB7KLOzZcSzQkCLleecU/\ntCclma1ZE5pzZmX5L5/u3c1iYvxfQEqK2fjxZr/8EppriEhklUGCcDwwB1gVWP4LnBDOa5ZmCdf9\nYvdus6uv9t+ff/qT2fbtJTtuyhSzhg39d++IEf77+FBycsyGDfPX+tvflBwEa+dOs1mzzB580KxH\nD7N69fYnDMccYzZggNlrr5lt2BCe6//vfz7Zi4311zzrLLO33zbLzjb77TezM84wq1bN7MMPw3N9\nkfyKu1eUupNypFTWTspF+eQT336/alU/tX3t2r4z1uG+/vQTvP46TJoE27b5UTb+/Ge/tGwZ6Z9S\nREKprOZBcM7VAjCzHeG+VmmE436xcaP/Tp450092OXr04XVA3rQJbr0VXn7Zz4Hz4ou+U21hzOCm\nm/zsxbfcAo8+quaeoZaT4yd/mz3bN02aPt2PmuScb5LUvbsfge+UU0rf6TknxzdFe+IJ37m6enW4\n9FI/qV3Boc23bIGzz4Zly/z9//TTCz1lubJvH/z3v/7v9cQT/RxH+jutGIq7VyhBqAAWLYKHH/Y3\nlu3bYceOA1937izZeWrWhIsv9v0Kzj47dKNqiEj5UgajGD0APGJmWwLrccAtZnZnuK5ZGqG+Xyxe\n7Dsjr1vn+xZceumhjynKJ5/4DscZGb7N+cMP+35oucx8UvD44z5JGDtWD11lIScHvvnGd3ieNs0P\ns5qT4+dgOO88nyycfz4cffShz/Xbb75v3xNPwPffwzHHwF//6vsZFnf8hg2+n8TGjT4RbdMmdD9f\nKH3/Pbz6qi98XLNm//batX2iUNiS/29cIk8JQiW3b9/BSUPB13r1/BjSNWtGOloRCbcySBDmm1n7\nAtu+MbMO4bpmaYTyfvGvf8GAAb5G9r33/Kg4wcqdWX7cOD+S3bPP+u9pM/jb3+Cxx+CGG/znSg4i\n49dffTKXmzBkZvp/i5QUnywUVruwciU89ZSvHdqyxe974437O7OXREaG70idk+PnSjruuLD8eIdt\n82Z4802fGMyZ4wsazz/fFzzWrw8//HDgkpHh/55zNWxYeOJw/PEaAj0SlCCIiESRMkgQFgGnmNnu\nwPrv8G1ZE8N1zdIIxf3CzD+o33YbdOgA777rm2iG0ldf+eGtlyyB/v19SfO4cTBsGDz5pJKD8uJQ\ntQtnn+2bEr3zjv83u+ginxiUdiTAJUvgzDN9Ad9//uP/LiIhO9sPI/vqq34Y2d27/TwTl18OAwcW\nP6P1rl2wfPnBicMPP/iaklzOQVyc/10ezlLeh64t75QgiIhEkTJIEG4D/gi8jB/yehDwvpk9Eq5r\nlkaw94vdu31zkFdf9aW/L78MsbEhDDCfPXvgoYfg/vth717f7OiZZ5QclGe//uoTgmnTfN+CzEz/\nkHvNNb4pUdOmhz7HoXz1FXTp4kvYv/jCJwtlZdEi/7efluYf5o880jerGzQI2rcP/m9zyxbf1+KH\nH+DHH33fj19/PXj57beizxEb6+M6+mjfT7NgzURZ/r4qIiUIIiJRpCw6KTvnugHnAoYf3e4YMxsW\nzmsermDuFxs2QO/e8OWXviPynXeWzcP6d9/5h8LLL1c/sYokJ8dPmJeQEPok8pNPoEcP36zt44/D\nl6SCf0ifMMEnBgsW+AFSLrzQ/z12717yJlKhtHt34YlD/uXnn32SsWKFb3adq0GDwps0nXACHHFE\n2f8s5U24ZlKWEkpL86NdrFoFzZrBmDG+Wk5EpALbgE8O+gIrgLciG07oLFjgOyNv3AiTJ/vBHcpK\nq1aadb4iqlIlfP9uXbv6h/Z+/fzf4rvvhvZB3QymTvUTs06d6psUJSf70bMGDIj8hHI1akDjxn45\nlD17Cm/S9OGHvgYwl3P+eSw3YWjUyCd5+/btXwquF7c0aOA7kyclQWJi5ejvqQQhzNLS/EgVWVl+\nfeVKvw5KEkSkYnHOnQgMCCwbgTfxNdHnRDSwEFq71o8gExfn2323b3/oY0TC7eKL4bnn/PPDoEHw\nxhvB1zBlZ/sOxw895EfoOuYYP2LW5Zf7PgYVUfXqcPLJfilo+/b9TZq+/35/8vD6637491zO+U7n\n+ZcqVQ7elv+z9ev3P+c555uEtWmzP2lo08ZvK+1QuZGgBCHMRo7c/0eTKyvLb1eCICIVzFJgFnCh\nmf0I4Jz7v8iGFFpNmvghRXv2jFynUJHCXH21H+58xAifwD71VOmavWVl+dL0xx7zowy1auWbFA0Y\nANWqhTzscqN2bT/QQIcCY62Z+ZqH3Af+0vxOc3J8zcW33/pl0SL/+t57/jPwHaoTEw9MGtq0KdmQ\nuZGgBCHMVq06vO0iIuVYH6A/MMM59yEwEd9JuVLJreUVKW9uu823uX/0Ud85d/Tokh+7ebPv+P7E\nE75D9e9/75sR9egR3f1dnAt+iNUqVXy/hhNO8H2XcmVl+X5F+ROHf//7wOZODRr45MXMJxOFvRb3\nWU6OH2Uq1JPqKUEIs2bNfLOiwraLiFQkZvYu8K5zribQC7gJONo59yzwjpl9HNEARaJA7sSp993n\nk4Qbbyx+/3Xr/IR7zz3n50bq3t3XQpxxhkbJCrfYWD8PRkqBbsAbNuxPGpYs8cPBVqni/z2Kei3u\ns3DUdipBCLMxYw7sgwD+D2bMmMjFJCISDDP7DZgATAjMotwXuA1QgiASZs75h/3Nm32fgbg4P1FZ\nQT/84GsaXnvN9zfo399PwNe2bdnHLAdq2NAv554b6UiKFsWVSmVj4EA/MkB8vP9PHR/v19X/QEQq\nAzPbbGapZtYl0rGIRIuqVf3IRl26wJVX+gnMcqWnQ9++vqPu66/7SfiWLfODpig5kJJSDUIZGDhQ\nCYGIiIiETo0aftbmLl38RH6PPeY7xU6fDnXq+GZEN97oS6pFDpcSBBEREZEKqHZtP5Nz585w/fW+\nLfrDD/vZnOvWjXR0UpEpQRARERGpoI48EmbMgJkz4Y9/1AzBEhpKEEREREQqsIYNfb8DkVBRJ2UR\nEREREcmjBEFERERERPIoQRARERERkTxKEEREpEw457o55753zv3onBtRyOfNnHMznHPznXOLnHPd\nA9sTnHM7nXMLAstzZR+9iEj0UCdlEREJO+dcDPA00BVYA8x1zr1vZt/l2+1OYJKZPeucawVMBRIC\nn/1kZu3KMmYRkWgV1hqEEpQW3eyc+y5QUvSpcy4+nPGIiEjEdAR+NLPlZrYHmAj0KrCPAXUC7+sC\n68owPhERCQhbgpCvtOgCoBUwIFAilN98IMXMkoApwCPhiqeiSkuDhASoUsW/pqVFOiIRkVJpAqzO\nt74msC2/e4DLnHNr8LUH1+f7rHmg6dEXzrnORV3EOTfEOZfunEvPzMwMUegiItElnDUIhywtMrMZ\nZpYVWJ0DHBvGeCqctDQYMgRWrgQz/zpkiJIEEam0BgCvmNmxQHfgdedcFWA90MzM2gM3AxOcc3UK\nO4GZpZpZipmlNGjQoMwCFxGpTMKZIJSktCi/q4BphX0QrSVCI0dCVtaB27Ky/HYRkQpmLdA03/qx\ngW35XQVMAjCzL4EjgKPMbLeZ/RrYPg/4CTgx7BGLiESpcjGKkXPuMiAFeLSwz6O1RGjVqsPbLiJS\njs0FWjjnmjvnqgP9gfcL7LMK6ALgnGuJTxAynXMNAs1Wcc4dB7QAlpdZ5CIiUSacCUJJSotwzp0L\njAR6mtnuMMZT4TRrdnjbRUTKKzPLBq4DPgL+hx+taIlzbrRzrmdgt1uAq51zC4F/AoPMzIAzgUXO\nuQX4/mrXmtmmsv8pRESiQziHOc0rLcInBv2BS/Pv4JxrDzwPdDOzX8IYS4U0Zozvc5C/mVFsrN8u\nIlLRmNlUfOfj/Nvuzvf+O6BTIce9BbwV9gBFRAQIYw1CCUuLHgVqAZMDk98UrG6OagMHQmoqxMeD\nc/41NdVvFxEREREJh7BOlFaC0qJzw3n9ymDgQCUEIiIiIlJ2ykUnZRERERERKR+UIIiIiIiISB4l\nCCIiIiIikkcJQiWXlgYJCVClin/VLMwiIiIiUpywdlKWyEpLO3CY1JUr/Tqo47OIiIiIFE41CJXY\nyJEHzqEAfn3kyMjEIyIiIiLlnxKESmzVqsPbLiIiIiKiBKESa9bs8LaLiIiIiChBqMTGjIHY2AO3\nxcb67SIiIiIihVGCUIkNHAipqRAfD87519RUdVAWERERkaJpFKNKbuBAJQQiIiIiUnKqQRARERER\nkTxKEKRYmmhNREREJLqoiZEUSROtiYiIiEQf1SBIkTTRmoiIiEj0UYIgRdJEayIiIiLRRwmCFEkT\nrYmIiIhEHyUIUiRNtCYiIiISfZQgSJE00ZqIiIhI9FGCIMUaOBAyMiAnx78ebnKgYVJFREREKhYN\ncypho2FSRURERCoe1SBI2GiYVBEREZGKJ6wJgnOum3Pue+fcj865EYV8fqZz7hvnXLZz7uJwxiJl\nT8OkioiIiFQ8YUsQnHMxwNPABUArYIBzrlWB3VYBg4AJ4YpDIkfDpIqIiIhUPOGsQegI/Ghmy81s\nDzAR6JV/BzPLMLNFQE4Y45AICXaYVHVwFhERESl74UwQmgCr862vCWw7bM65Ic65dOdcemZmZkiC\nk/ALZpjU3A7OK1eC2f4OzkoSRERERMKrQnRSNrNUM0sxs5QGDRpEOhw5DKUdJlUdnEVEREQiI5wJ\nwlqgab71YwPbRA5JHZxFREREIiOcCcJcoIVzrrlzrjrQH3g/jNeTSkQdnEVEREQiI2wJgpllA9cB\nHwH/AyaZ2RLn3GjnXE8A59wpzrk1QF/geefcknDFIxVLsB2cQZ2cRUREREojrDMpm9lUYGqBbXfn\nez8X3/RI5AC5fRVGjvTNipo188lBSfswaBZnERERkdKpEJ2UJTqVtoMzqJOziIiISGkpQZBKSZ2c\nRUREyo9QN/vdsmULzzzzTKmO7d69O1u2bCl2n7vvvpvp06eX6vyVgRIEqZRC0clZfRhERESCF465\njYpLELKzs4s9durUqdSrV6/YfUaPHs25555b6vgqOiUIUimFYhZnTdQmElrOuW7Oue+dcz8650YU\n8nkz59wM59x859wi51z3fJ/dHjjue+fc+WUbuYgEIxzNfkeMGMFPP/1Eu3btGD58OJ9//jmdO3em\nZ8+etGrVCoA//elPJCcnk5iYSGpqat6xCQkJbNy4kYyMDFq2bMnVV19NYmIi5513Hjt37gRg0KBB\nTJkyJW//UaNG0aFDB9q0acPSpUsByMzMpGvXriQmJjJ48GDi4+PZuHHjQbEOHTqUlJQUEhMTGTVq\nVN72uXPncvrpp9O2bVs6duzI9u3b2bdvH7feeiutW7cmKSmJJ598svS/pGCYWYVakpOTTaQk3njD\nLD7ezDn/+sYbJT82Pt7MpwYHLvHx4YlVJJSAdCsH39f5FyAG+Ak4DqgOLARaFdgnFRgaeN8KyMj3\nfiFQA2geOE/Moa6p+4VI+eBc4fdU50p/zhUrVlhiYmLe+owZMyw2NtaWL1+et+3XX381M7OsrCxL\nTEy0jRs3mplZfHy8ZWZm2ooVKywmJsbmz59vZmZ9+/a1119/3czMLr/8cps8eXLe/uPHjzczs6ef\nftquuuoqMzMbNmyYPfDAA2ZmNm3aNAMsMzPzoFhz48jOzrazzjrLFi5caLt377bmzZvb119/bWZm\nW7dutb1799ozzzxjF110ke3du/eAY8OhuHtFWEcxEomkgQNLP2KR+jCIhFxH4EczWw7gnJsI9AK+\ny7ePAXUC7+sC6wLvewETzWw3sMI592PgfF+WReAiEpxmzXxNfGHbQ6ljx440b948b338+PG88847\nAKxevZply5Zx5JFHHnBM8+bNadeuHQDJyclkZGQUeu4+ffrk7fP2228DMHv27Lzzd+vWjbi4uEKP\nnTRpEqmpqWRnZ7N+/Xq+++47nHM0atSIU045BYA6dfxX3/Tp07n22mupWtU/otevX/+wfw+hoCZG\nIoVQHwaRkGsCrM63viawLb97gMsC8+NMBa4/jGMBcM4Ncc6lO+fSMzMzQxG3iAQpFHMblUTNmjXz\n3n/++edMnz6dL7/8koULF9K+fXt27dp10DE1atTIex8TE1Nk/4Xc/YrbpzArVqzgscce49NPP2XR\nokX06NGj0DjKGyUIIoVQHwaRiBgAvGJmxwLdgdedc4d1nzKzVDNLMbOUBg0ahCVIETk8AwdCairE\nx4Nz/jU1Nbh5iWrXrs327duL/Hzr1q3ExcURGxvL0qVLmTNnTukvVoROnToxadIkAD7++GM2b958\n0D7btm2jZs2a1K1blw0bNjBt2jQATjrpJNavX8/cuXMB2L59O9nZ2XTt2pXnn38+LwnZtGlTyOMu\nCSUIIoUI9sssFB2yVAMhlcxaoGm+9WMD2/K7CpgEYGZfAkcAR5XwWBEpx4KZ26gwRx55JJ06daJ1\n69YMHz78oM+7detGdnY2LVu2ZMSIEZx22mnBXbAQo0aN4uOPP6Z169ZMnjyZY445htq1ax+wT9u2\nbWnfvj0nn3wyl156KZ06dQKgevXqvPnmm1x//fW0bduWrl27smvXLgYPHkyzZs1ISkqibdu2TJgw\nAfDDrr7//vsh/xmK4nwfhYojJSXF0tPTIx2GSLGqVPE1BwU5578cD6XgTNDgazCCLXGR6OCcm2dm\nKZGOIz/nXFXgB6AL/uF+LnCpmS3Jt8804E0ze8U51xL4FN+UqBUwAd/voHFgewsz21fcNXW/EJFw\n2r17NzExMVStWpUvv/ySoUOHsmDBgkiHVWLF3SvUSVkkDILtkFVcDYQSBKmIzCzbOXcd8BF+RKOX\nzGyJc240fiSN94FbgBecc/+H77A8KDDSxhLn3CR8h+ZsYNihkgMRkXBbtWoVl1xyCTk5OVSvXp0X\nXngh0iGFjGoQRMIg2BqAUNRAjBzpR11q1sz3nVBiET3KYw1CJOh+ISJStOLuFeqDIBIGwfZhCGYU\nJXWQFhERkWAoQRAJk2A6ZAUzipI6SIuIiEgwlCCIlEPB1EAEO8mbaiBERESimxIEkXKqtDUQwU7y\nphoIERGR6KYEQaSSCXaSt/JQA6EEQ0REQq1WrVoArFu3josvvrjQfc4++2wONbjBuHHjyMpXkta9\ne3e2bNkSukDLASUIIpVMJDtIQ/A1EGriJCIi4dS4cWOmTJlS6uMLJghTp06lXr16oQit3NA8CCKV\n0MCBpR/WdMyYwodoFlY69QAADp1JREFULasaCM0BISJSsdx0E4R6frB27WDcuKI/HzFiBE2bNmXY\nsGEA3HPPPdSqVYtrr72WXr16sXnzZvbu3cv9999Pr169Djg2IyODCy+8kMWLF7Nz506uuOIKFi5c\nyMknn8zOnTvz9hs6dChz585l586dXHzxxdx7772MHz+edevWcc4553DUUUcxY8YMEhISSE9P56ij\njmLs2LG89NJLAAwePJibbrqJjIwMLrjgAs444wz++9//0qRJE9577z1+97vfHRDXBx98wP3338+e\nPXs48sgjSUtLo2HDhuzYsYPrr7+e9PR0nHOMGjWKiy66iA8//JA77riDffv2cdRRR/Hpp5+G6Lev\nGgQRKSDSNRChaOIUTPMkNW8SESn/+vXrx6RJk/LWJ02aRL9+/TjiiCN45513+Oabb5gxYwa33HIL\nxc359eyzzxIbG8v//vc/7r33XubNm5f32ZgxY0hPT2fRokV88cUXLFq0iBtuuIHGjRszY8YMZsyY\nccC55s2bx8svv8xXX33FnDlzeOGFF5g/fz4Ay5YtY9iwYSxZsoR69erx1ltvHRTLGWecwZw5c5g/\nfz79+/fnkUceAeC+++6jbt26fPvttyxatIg//OEPZGZmcvXVV/PWW2+xcOFCJk+eHNTvsyDVIIjI\nQSJZAxHMLNQFJ6jLbZ4EJft5gj0+9xyapE5EoklxJf3h0r59e3755RfWrVtHZmYmcXFxNG3alL17\n93LHHXcwc+ZMqlSpwtq1a9mwYQPHHHNMoeeZOXMmN9xwAwBJSUkkJSXlfTZp0iRSU1PJzs5m/fr1\nfPfddwd8XtDs2bPp3bs3NWvWBKBPnz7MmjWLnj170rx5c9q1awdAcnIyGRkZBx2/Zs0a+vXrx/r1\n69mzZw/NmzcHYPr06UycODFvv7i4OD744APOPPPMvH3q169/GL+9Q1MNgoiEVLA1EJGcA6I89J9Q\nDUZk6fcvUnH07duXKVOm8Oabb9KvXz8A0tLSyMzMZN68eSxYsICGDRuya9euwz73ihUreOyxx/j0\n009ZtGgRPXr0KNV5ctWoUSPvfUxMDNnZ2Qftc/3113Pdddfx7bff8vzzzwd1vWCFNUFwznVzzn3v\nnPvROTeikM9rOOfeDHz+lXMuIZzxiEjZCGaSuEjOARHO/hMloQ7akaXfv0jF0q9fPyZOnMiUKVPo\n27cvAFu3buXoo4+mWrVqzJgxg5WFVUnnc+aZZzJhwgQAFi9ezKJF/9/e/cdIUd5xHH9/BEwMEkQh\nlh5UtJKiSM87Log/MNpWRdpoS0tFSAoqIZLa6h9tpdEQ0jRptOmPaEkbrFawppqmhZIGVKqkNKla\nTwMoagsqiRJAlJMf0lqx3/4xw7quu8fe7u3O3N3nlUxu5plnZr/77HPzzbMzs7MFgAMHDjB06FCG\nDx/Onj17WLduXWGbYcOGcfDgwY/ta9q0aaxevZrDhw/z7rvvsmrVKqZNm1b1+9m/fz8tLS0ArFix\nolB+2WWXsWzZssJyV1cXU6dOZePGjbz22msA7Nu3r+rXqUbDBgiSBgHLgCuBs4FrJZ1dUu0GoCsi\nzgR+BtzRqHjMrO/I6hkQWd8/0RvPoLDauf3N+paJEydy8OBBWlpaGD16NABz586ls7OTSZMmsXLl\nSiZMmNDtPhYtWsShQ4c466yzWLJkCZMnTwagtbWVtrY2JkyYwJw5c7jwwgsL2yxcuJDp06dz6aWX\nfmRf7e3tzJ8/nylTpnDeeeexYMEC2traqn4/S5cuZdasWUyePJmRI0cWym+//Xa6uro455xzaG1t\nZcOGDYwaNYrly5czc+ZMWltbC2dQOjs7WbBgQdWvWYm6u3Gjrh1L5wNLI+KKdPn7ABHxo6I6j6Z1\nnpQ0GNgNjIpuguro6Ihj/T6tmQ1MpfcQQHJ5UrVnIOrdfty48vdPnHZaMtA5luOOS765LiUlg6Vq\nSXo2Ijqq36J/6mm+6K32NzPrC7rLFY28xKgFeL1o+Y20rGydiDgC7AdOKd2RpIWSOiV17t27t0Hh\nmllfV+/9D1nePwH1n8Gw+rj9zcwSfeIm5YhYHhEdEdExatSorMMxsxyr5/6HerfPeoBh9XH7m5kl\nGjlA2AmMLVoek5aVrZNeYjQceLuBMZmZNVSWAwyrj9vfzCzRyOcgPAOMl3Q6yUBgNjCnpM4aYB7w\nJPA14Inu7j8wM+vv6nkGhdXP7W9m1sABQkQckXQT8CgwCLgvIrZK+gHQGRFrgHuBByRtB/aRDCLM\nzMzMzCwjDX2SckSsBdaWlC0pmv8PMKuRMZiZmZmZWfX6xE3KZmZmZmbWHB4gmJmZmZlZgQcIZmZm\nZmZW0LAnKTeKpL1AmWeV5sJI4K2sg6ggz7FBvuNzbLXLc3x5jg3qi++0iBjwD43Jcb7oz32v0fIc\nG+Q7vjzHBvmOr7/GVjFX9LkBQp5J6qz0yOqs5Tk2yHd8jq12eY4vz7FB/uOz2uX9s81zfHmODfId\nX55jg3zHNxBj8yVGZmZmZmZW4AGCmZmZmZkVeIDQu5ZnHUA38hwb5Ds+x1a7PMeX59gg//FZ7fL+\n2eY5vjzHBvmOL8+xQb7jG3Cx+R4EMzMzMzMr8BkEMzMzMzMr8ADBzMzMzMwKPEDoAUljJW2Q9KKk\nrZJuLlPnEkn7JW1KpyVNjnGHpOfT1+4ss16S7pK0XdIWSe1NiuszRW2ySdIBSbeU1Glq20m6T9Kb\nkl4oKjtZ0npJ29K/IypsOy+ts03SvCbF9mNJL6ef2ypJJ1XYtts+0MD4lkraWfT5zaiw7XRJ/0z7\n4OImxfZwUVw7JG2qsG0z2q7scSQvfc96h/NFXXE5X9Qfm/NF7bHlIl9knisiwlOVEzAaaE/nhwH/\nAs4uqXMJ8OcMY9wBjOxm/QxgHSBgKvB0BjEOAnaTPKAjs7YDLgbagReKyu4EFqfzi4E7ymx3MvBq\n+ndEOj+iCbFdDgxO5+8oF1s1faCB8S0FvlPFZ/8KcAZwPLC59H+oEbGVrP8JsCTDtit7HMlL3/PU\n2M+5pI7zxbFjdL6oLTbnixpjK1mfWb7IOlf4DEIPRMSuiHgunT8IvAS0ZBtVj10NrIzEU8BJkkY3\nOYbPA69ERKZPOI2IjcC+kuKrgRXp/Argy2U2vQJYHxH7IqILWA9Mb3RsEfFYRBxJF58CxvTma/ZE\nhbarxhRge0S8GhH/BR4iafOmxCZJwNeB3/Xma/ZEN8eRXPQ96x3OF73G+aKG2Jwv6o8t63yRda7w\nAKFGksYBbcDTZVafL2mzpHWSJjY1MAjgMUnPSlpYZn0L8HrR8hs0P2nNpvI/XJZtB3BqROxK53cD\np5apk4c2vJ7km71yjtUHGumm9JT2fRVOe2bddtOAPRGxrcL6prZdyXGkr/Q96yHni7o4X9TP+aI2\nuckXWeQKDxBqIOlE4A/ALRFxoGT1cySnQluBu4HVTQ7voohoB64Evinp4ia/frckHQ9cBfy+zOqs\n2+4jIjlPl7vfAZZ0G3AEeLBClaz6wC+BTwPnArtITs3mzbV0/21Q09quu+NIXvue9ZzzRe2cL+rn\nfFGXXOSLrHKFBwg9JGkIyQf1YET8sXR9RByIiEPp/FpgiKSRzYovInamf98EVpGcoiu2ExhbtDwm\nLWuWK4HnImJP6Yqs2y615+gp9PTvm2XqZNaGkuYDXwLmpgeGj6miDzREROyJiA8i4n/APRVeN8u2\nGwzMBB6uVKdZbVfhOJLrvmc953xRN+eLOjhf1C4v+SLLXOEBQg+k16PdC7wUET+tUOcTaT0kTSFp\n47ebFN9QScOOzpPcpPRCSbU1wDeUmArsLzpV1QwVR+RZtl2RNcDRu/3nAX8qU+dR4HJJI9LTopen\nZQ0laTrwPeCqiDhcoU41faBR8RVfm/yVCq/7DDBe0unpt4OzSdq8Gb4AvBwRb5Rb2ay26+Y4ktu+\nZz3nfNErnC9q5HxRt8zzRea5Ihp453p/m4CLSE7lbAE2pdMM4EbgxrTOTcBWkrvtnwIuaGJ8Z6Sv\nuzmN4ba0vDg+ActIfhngeaCjifENJTmADy8qy6ztSBLPLuB9kuvzbgBOAR4HtgF/AU5O63YAvy7a\n9npgezpd16TYtpNcU3i07/0qrftJYG13faBJ8T2Q9qktJAew0aXxpcszSH6N4ZVGxFcutrT8/qN9\nrahuFm1X6TiSi77nqeGfs/NFdfE5X9QXm/NFjbGl5feTcb7o5hjSlH6ndCdmZmZmZma+xMjMzMzM\nzD7kAYKZmZmZmRV4gGBmZmZmZgUeIJiZmZmZWYEHCGZmZmZmVuABgg1okj6QtKloWtyL+x4nqSm/\nKW1mZo3lfGEDyeCsAzDL2L8j4tysgzAzs9xzvrABw2cQzMqQtEPSnZKel/QPSWem5eMkPSFpi6TH\nJX0qLT9V0ipJm9PpgnRXgyTdI2mrpMcknZDW/7akF9P9PJTR2zQzszo5X1h/5AGCDXQnlJwyvqZo\n3f6ImAT8Avh5WnY3sCIiPgs8CNyVlt8F/DUiWoF2kicrAowHlkXEROAd4Ktp+WKgLd3PjY16c2Zm\n1mucL2zA8JOUbUCTdCgiTixTvgP4XES8KmkIsDsiTpH0Fskj4d9Py3dFxEhJe4ExEfFe0T7GAesj\nYny6fCswJCJ+KOkR4BCwGlgdEYca/FbNzKwOzhc2kPgMglllUWG+J94rmv+AD+/7+SKwjOTbo2ck\n+X4gM7O+y/nC+hUPEMwqu6bo75Pp/N+B2en8XOBv6fzjwCIASYMkDa+0U0nHAWMjYgNwKzAc+Ni3\nUmZm1mc4X1i/4lGoDXQnSNpUtPxIRBz96boRkraQfKtzbVr2LeA3kr4L7AWuS8tvBpZLuoHkm59F\nwK4KrzkI+G2aFATcFRHv9No7MjOzRnC+sAHD9yCYlZFeU9oREW9lHYuZmeWX84X1R77EyMzMzMzM\nCnwGwczMzMzMCnwGwczMzMzMCjxAMDMzMzOzAg8QzMzMzMyswAMEMzMzMzMr8ADBzMzMzMwK/g+R\nfVdBC8AuygAAAABJRU5ErkJggg==\n","text/plain":["<Figure size 936x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"markdown","metadata":{"id":"8uX3N1mjeSm8","colab_type":"text"},"source":["In order to avoid overfitting, the model will be trained over 4 epochs."]},{"cell_type":"code","metadata":{"id":"EOYsuUBZdhgB","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":170},"outputId":"061fdf4b-8336-4576-ba4d-b63d6c89daba","executionInfo":{"status":"ok","timestamp":1577083433723,"user_tz":-60,"elapsed":9866,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer=optimizers.RMSprop(lr=0.001),\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=4,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":57,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/4\n","15000/15000 [==============================] - 3s 183us/step - loss: 0.5495 - acc: 0.7516 - val_loss: 0.4224 - val_acc: 0.8535\n","Epoch 2/4\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.3335 - acc: 0.8987 - val_loss: 0.3206 - val_acc: 0.8890\n","Epoch 3/4\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.2398 - acc: 0.9261 - val_loss: 0.2852 - val_acc: 0.8914\n","Epoch 4/4\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.1801 - acc: 0.9453 - val_loss: 0.2834 - val_acc: 0.8871\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"eMQVsUraeReT","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":51},"outputId":"551264fe-d4aa-496e-eb94-4d0777f12213","executionInfo":{"status":"ok","timestamp":1577083441406,"user_tz":-60,"elapsed":5122,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["results = model.evaluate(x_test,y_test)\n","results"],"execution_count":58,"outputs":[{"output_type":"stream","text":["25000/25000 [==============================] - 5s 186us/step\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["[0.2999872021770477, 0.88012]"]},"metadata":{"tags":[]},"execution_count":58}]},{"cell_type":"markdown","metadata":{"id":"e_N17ahNeqad","colab_type":"text"},"source":["## Neural network 3\n","\n","- 3 hidden layers (instead of 2)\n","- number of units: 16\n","- loss function: binary_crossentropy\n","- activation function: relu"]},{"cell_type":"code","metadata":{"id":"Vgv3XkHHebff","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":714},"outputId":"f8a72e4f-cdf6-44a9-c0ca-b108646d85d8","executionInfo":{"status":"ok","timestamp":1577083489940,"user_tz":-60,"elapsed":43160,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer='rmsprop',\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=20,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":59,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/20\n","15000/15000 [==============================] - 3s 194us/step - loss: 0.5349 - acc: 0.7611 - val_loss: 0.4059 - val_acc: 0.8479\n","Epoch 2/20\n","15000/15000 [==============================] - 2s 142us/step - loss: 0.2958 - acc: 0.9015 - val_loss: 0.3085 - val_acc: 0.8763\n","Epoch 3/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.2167 - acc: 0.9252 - val_loss: 0.2916 - val_acc: 0.8824\n","Epoch 4/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.1640 - acc: 0.9463 - val_loss: 0.2791 - val_acc: 0.8872\n","Epoch 5/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.1313 - acc: 0.9553 - val_loss: 0.2922 - val_acc: 0.8873\n","Epoch 6/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.1076 - acc: 0.9647 - val_loss: 0.3305 - val_acc: 0.8801\n","Epoch 7/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0833 - acc: 0.9741 - val_loss: 0.3975 - val_acc: 0.8629\n","Epoch 8/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0687 - acc: 0.9795 - val_loss: 0.3767 - val_acc: 0.8734\n","Epoch 9/20\n","15000/15000 [==============================] - 2s 136us/step - loss: 0.0534 - acc: 0.9845 - val_loss: 0.4573 - val_acc: 0.8607\n","Epoch 10/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0402 - acc: 0.9901 - val_loss: 0.4713 - val_acc: 0.8682\n","Epoch 11/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0338 - acc: 0.9915 - val_loss: 0.6699 - val_acc: 0.8394\n","Epoch 12/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.0333 - acc: 0.9909 - val_loss: 0.4885 - val_acc: 0.8714\n","Epoch 13/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0124 - acc: 0.9985 - val_loss: 0.5346 - val_acc: 0.8701\n","Epoch 14/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0177 - acc: 0.9961 - val_loss: 0.5458 - val_acc: 0.8717\n","Epoch 15/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0170 - acc: 0.9958 - val_loss: 0.5790 - val_acc: 0.8690\n","Epoch 16/20\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.0049 - acc: 0.9997 - val_loss: 0.6107 - val_acc: 0.8685\n","Epoch 17/20\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.0156 - acc: 0.9949 - val_loss: 0.6419 - val_acc: 0.8685\n","Epoch 18/20\n","15000/15000 [==============================] - 2s 135us/step - loss: 0.0026 - acc: 0.9999 - val_loss: 0.6699 - val_acc: 0.8675\n","Epoch 19/20\n","15000/15000 [==============================] - 2s 135us/step - loss: 0.0158 - acc: 0.9953 - val_loss: 0.7429 - val_acc: 0.8621\n","Epoch 20/20\n","15000/15000 [==============================] - 2s 136us/step - loss: 0.0016 - acc: 0.9999 - val_loss: 0.7308 - val_acc: 0.8669\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"HQB9k9JPeuDG","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":279},"outputId":"0ac90aaf-e816-4328-a33e-7eadbdc3ec79","executionInfo":{"status":"ok","timestamp":1577083529341,"user_tz":-60,"elapsed":1001,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["history_dict= history.history\n","visualize_history(history_dict)"],"execution_count":60,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAwgAAAEGCAYAAADBkLJOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niIjEqYEgIiIiIiJx/w8ukqkAhbw8AQAAAABJRU5ErkJggg==\n","text/plain":["<Figure size 936x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"markdown","metadata":{"id":"k2p9pCBRgc89","colab_type":"text"},"source":["In order to avoid overfitting, the model will be trained over 4 epochs."]},{"cell_type":"code","metadata":{"id":"5Dz9qMMXgORD","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":170},"outputId":"66a1a933-085e-44db-e709-74dd39b339fe","executionInfo":{"status":"ok","timestamp":1577083583537,"user_tz":-60,"elapsed":10328,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer=optimizers.RMSprop(lr=0.001),\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=4,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":61,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/4\n","15000/15000 [==============================] - 3s 194us/step - loss: 0.5459 - acc: 0.7707 - val_loss: 0.4017 - val_acc: 0.8684\n","Epoch 2/4\n","15000/15000 [==============================] - 2s 138us/step - loss: 0.3142 - acc: 0.9017 - val_loss: 0.3090 - val_acc: 0.8824\n","Epoch 3/4\n","15000/15000 [==============================] - 2s 137us/step - loss: 0.2246 - acc: 0.9239 - val_loss: 0.2853 - val_acc: 0.8863\n","Epoch 4/4\n","15000/15000 [==============================] - 2s 136us/step - loss: 0.1730 - acc: 0.9411 - val_loss: 0.2747 - val_acc: 0.8926\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"z4VNuPesgZOF","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":51},"outputId":"4b0cc912-88d0-484c-f066-25c869c7c652","executionInfo":{"status":"ok","timestamp":1577083604119,"user_tz":-60,"elapsed":4899,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["results = model.evaluate(x_test,y_test)\n","results"],"execution_count":62,"outputs":[{"output_type":"stream","text":["25000/25000 [==============================] - 5s 183us/step\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["[0.29519664472579954, 0.88248]"]},"metadata":{"tags":[]},"execution_count":62}]},{"cell_type":"markdown","metadata":{"id":"E_lq7kM-g7b0","colab_type":"text"},"source":["## Neural network 4\n","\n","- 2 hidden layers \n","- number of units: 64 (instead of 16)\n","- loss function: binary_crossentropy\n","- activation function: relu"]},{"cell_type":"code","metadata":{"id":"vVh8U1Vxgfkh","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":714},"outputId":"d73f71f0-1503-4f76-af61-be03065c030b","executionInfo":{"status":"ok","timestamp":1577083770013,"user_tz":-60,"elapsed":59622,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(64,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer='rmsprop',\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=20,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":63,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/20\n","15000/15000 [==============================] - 4s 254us/step - loss: 0.4765 - acc: 0.7894 - val_loss: 0.3201 - val_acc: 0.8796\n","Epoch 2/20\n","15000/15000 [==============================] - 3s 191us/step - loss: 0.2654 - acc: 0.9003 - val_loss: 0.2813 - val_acc: 0.8859\n","Epoch 3/20\n","15000/15000 [==============================] - 3s 192us/step - loss: 0.1902 - acc: 0.9298 - val_loss: 0.3085 - val_acc: 0.8748\n","Epoch 4/20\n","15000/15000 [==============================] - 3s 193us/step - loss: 0.1462 - acc: 0.9461 - val_loss: 0.2908 - val_acc: 0.8879\n","Epoch 5/20\n","15000/15000 [==============================] - 3s 194us/step - loss: 0.1148 - acc: 0.9580 - val_loss: 0.3561 - val_acc: 0.8750\n","Epoch 6/20\n","15000/15000 [==============================] - 3s 194us/step - loss: 0.0879 - acc: 0.9698 - val_loss: 0.3447 - val_acc: 0.8824\n","Epoch 7/20\n","15000/15000 [==============================] - 3s 196us/step - loss: 0.0730 - acc: 0.9758 - val_loss: 0.3794 - val_acc: 0.8802\n","Epoch 8/20\n","15000/15000 [==============================] - 3s 193us/step - loss: 0.0557 - acc: 0.9811 - val_loss: 0.4131 - val_acc: 0.8781\n","Epoch 9/20\n","15000/15000 [==============================] - 3s 194us/step - loss: 0.0543 - acc: 0.9852 - val_loss: 0.4507 - val_acc: 0.8764\n","Epoch 10/20\n","15000/15000 [==============================] - 3s 192us/step - loss: 0.0146 - acc: 0.9978 - val_loss: 0.5176 - val_acc: 0.8680\n","Epoch 11/20\n","15000/15000 [==============================] - 3s 193us/step - loss: 0.0445 - acc: 0.9893 - val_loss: 0.5344 - val_acc: 0.8746\n","Epoch 12/20\n","15000/15000 [==============================] - 3s 190us/step - loss: 0.0064 - acc: 0.9995 - val_loss: 0.5995 - val_acc: 0.8662\n","Epoch 13/20\n","15000/15000 [==============================] - 3s 192us/step - loss: 0.0406 - acc: 0.9898 - val_loss: 0.6137 - val_acc: 0.8711\n","Epoch 14/20\n","15000/15000 [==============================] - 3s 192us/step - loss: 0.0028 - acc: 0.9999 - val_loss: 0.6471 - val_acc: 0.8696\n","Epoch 15/20\n","15000/15000 [==============================] - 3s 192us/step - loss: 0.0048 - acc: 0.9992 - val_loss: 1.1756 - val_acc: 0.8107\n","Epoch 16/20\n","15000/15000 [==============================] - 3s 191us/step - loss: 0.0075 - acc: 0.9981 - val_loss: 0.7240 - val_acc: 0.8688\n","Epoch 17/20\n","15000/15000 [==============================] - 3s 191us/step - loss: 9.0710e-04 - acc: 0.9999 - val_loss: 0.7798 - val_acc: 0.8693\n","Epoch 18/20\n","15000/15000 [==============================] - 3s 193us/step - loss: 0.0332 - acc: 0.9925 - val_loss: 0.8111 - val_acc: 0.8679\n","Epoch 19/20\n","15000/15000 [==============================] - 3s 194us/step - loss: 4.2126e-04 - acc: 1.0000 - val_loss: 0.8302 - val_acc: 0.8686\n","Epoch 20/20\n","15000/15000 [==============================] - 3s 192us/step - loss: 3.0581e-04 - acc: 1.0000 - val_loss: 0.8593 - val_acc: 0.8681\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"sampKVsWg6tg","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":279},"outputId":"06fd2dc0-a6b8-410c-87f1-2fd4414d0eef","executionInfo":{"status":"ok","timestamp":1577083842781,"user_tz":-60,"elapsed":961,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["history_dict= history.history\n","visualize_history(history_dict)"],"execution_count":64,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAwgAAAEGCAYAAADBkLJOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n","text/plain":["<Figure size 936x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"markdown","metadata":{"id":"hFIaWpZZhw4G","colab_type":"text"},"source":["In order to avoid overfitting, the model will be trained over 2 epochs."]},{"cell_type":"code","metadata":{"id":"OId8n4YbhazD","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":102},"outputId":"3831de0d-5924-4706-b07e-5f424ff6229a","executionInfo":{"status":"ok","timestamp":1577083878448,"user_tz":-60,"elapsed":8278,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(64,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer='rmsprop',\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=2,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":65,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/2\n","15000/15000 [==============================] - 4s 268us/step - loss: 0.4855 - acc: 0.7721 - val_loss: 0.3472 - val_acc: 0.8583\n","Epoch 2/2\n","15000/15000 [==============================] - 3s 196us/step - loss: 0.2611 - acc: 0.9051 - val_loss: 0.2756 - val_acc: 0.8914\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"naxZaAzhhhuC","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":51},"outputId":"f1ef3bb1-4152-48be-8278-7991fe6c79e3","executionInfo":{"status":"ok","timestamp":1577083891667,"user_tz":-60,"elapsed":2773,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["results = model.evaluate(x_test,y_test)\n","results"],"execution_count":66,"outputs":[{"output_type":"stream","text":["25000/25000 [==============================] - 2s 92us/step\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["[0.2910332079124451, 0.88572]"]},"metadata":{"tags":[]},"execution_count":66}]},{"cell_type":"markdown","metadata":{"id":"-hmrstLBh8oP","colab_type":"text"},"source":["## Neural network 5\n","\n","- 2 hidden layers \n","- number of units: 16\n","- loss function: mse\n","- activation function: relu"]},{"cell_type":"code","metadata":{"id":"7f8MdIqshmQX","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":714},"outputId":"8c404c6a-cd55-4a08-9543-047de73632b6","executionInfo":{"status":"ok","timestamp":1577084011058,"user_tz":-60,"elapsed":45054,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer='rmsprop',\n","                       loss='mse',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=20,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":68,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/20\n","15000/15000 [==============================] - 3s 217us/step - loss: 0.1762 - acc: 0.7822 - val_loss: 0.1254 - val_acc: 0.8680\n","Epoch 2/20\n","15000/15000 [==============================] - 2s 145us/step - loss: 0.0968 - acc: 0.9018 - val_loss: 0.0965 - val_acc: 0.8857\n","Epoch 3/20\n","15000/15000 [==============================] - 2s 143us/step - loss: 0.0686 - acc: 0.9289 - val_loss: 0.0890 - val_acc: 0.8852\n","Epoch 4/20\n","15000/15000 [==============================] - 2s 145us/step - loss: 0.0529 - acc: 0.9431 - val_loss: 0.0842 - val_acc: 0.8891\n","Epoch 5/20\n","15000/15000 [==============================] - 2s 144us/step - loss: 0.0431 - acc: 0.9538 - val_loss: 0.0868 - val_acc: 0.8832\n","Epoch 6/20\n","15000/15000 [==============================] - 2s 143us/step - loss: 0.0360 - acc: 0.9635 - val_loss: 0.0833 - val_acc: 0.8861\n","Epoch 7/20\n","15000/15000 [==============================] - 2s 144us/step - loss: 0.0294 - acc: 0.9713 - val_loss: 0.0855 - val_acc: 0.8846\n","Epoch 8/20\n","15000/15000 [==============================] - 2s 146us/step - loss: 0.0245 - acc: 0.9779 - val_loss: 0.0950 - val_acc: 0.8718\n","Epoch 9/20\n","15000/15000 [==============================] - 2s 145us/step - loss: 0.0213 - acc: 0.9819 - val_loss: 0.0889 - val_acc: 0.8794\n","Epoch 10/20\n","15000/15000 [==============================] - 2s 146us/step - loss: 0.0174 - acc: 0.9857 - val_loss: 0.0904 - val_acc: 0.8772\n","Epoch 11/20\n","15000/15000 [==============================] - 2s 146us/step - loss: 0.0153 - acc: 0.9871 - val_loss: 0.0928 - val_acc: 0.8779\n","Epoch 12/20\n","15000/15000 [==============================] - 2s 142us/step - loss: 0.0128 - acc: 0.9898 - val_loss: 0.0956 - val_acc: 0.8743\n","Epoch 13/20\n","15000/15000 [==============================] - 2s 143us/step - loss: 0.0100 - acc: 0.9930 - val_loss: 0.0973 - val_acc: 0.8722\n","Epoch 14/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0089 - acc: 0.9934 - val_loss: 0.1000 - val_acc: 0.8714\n","Epoch 15/20\n","15000/15000 [==============================] - 2s 144us/step - loss: 0.0076 - acc: 0.9944 - val_loss: 0.1019 - val_acc: 0.8696\n","Epoch 16/20\n","15000/15000 [==============================] - 2s 143us/step - loss: 0.0068 - acc: 0.9953 - val_loss: 0.1033 - val_acc: 0.8686\n","Epoch 17/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0062 - acc: 0.9952 - val_loss: 0.1050 - val_acc: 0.8676\n","Epoch 18/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0045 - acc: 0.9966 - val_loss: 0.1073 - val_acc: 0.8670\n","Epoch 19/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0053 - acc: 0.9955 - val_loss: 0.1087 - val_acc: 0.8638\n","Epoch 20/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0047 - acc: 0.9958 - val_loss: 0.1097 - val_acc: 0.8633\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"QfTLZB5Jh3La","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":279},"outputId":"3b4d4d85-1c52-43c4-8247-05df0f8f31f8","executionInfo":{"status":"ok","timestamp":1577084021482,"user_tz":-60,"elapsed":896,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["history_dict= history.history\n","visualize_history(history_dict)"],"execution_count":69,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAxUAAAEGCAYAAAATwLgWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmVnjMCNBSv8AAAAASUVORK5CYII=\n","text/plain":["<Figure size 936x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"id":"DKCxZLWniGbs","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":170},"outputId":"5a7f7ce6-0e8b-40f7-d382-d279be825d45","executionInfo":{"status":"ok","timestamp":1577084061953,"user_tz":-60,"elapsed":10222,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer='rmsprop',\n","                       loss='mse',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=4,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":70,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/4\n","15000/15000 [==============================] - 3s 213us/step - loss: 0.1761 - acc: 0.7859 - val_loss: 0.1362 - val_acc: 0.8371\n","Epoch 2/4\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0969 - acc: 0.9021 - val_loss: 0.0961 - val_acc: 0.8865\n","Epoch 3/4\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0698 - acc: 0.9265 - val_loss: 0.0871 - val_acc: 0.8894\n","Epoch 4/4\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0538 - acc: 0.9449 - val_loss: 0.0845 - val_acc: 0.8877\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"qQhdo5WZiOC2","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":51},"outputId":"a68505e3-0b52-42c5-f234-87f5677854e3","executionInfo":{"status":"ok","timestamp":1577084067994,"user_tz":-60,"elapsed":4925,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["results = model.evaluate(x_test,y_test)\n","results"],"execution_count":71,"outputs":[{"output_type":"stream","text":["25000/25000 [==============================] - 4s 177us/step\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["[0.08975885879158974, 0.88148]"]},"metadata":{"tags":[]},"execution_count":71}]},{"cell_type":"markdown","metadata":{"id":"l6tjEFeulDkq","colab_type":"text"},"source":["## Neural network 6\n","\n","- 2 hidden layers \n","- number of units: 16\n","- loss function: binary_crossentropy\n","- activation function: tanh (instead of relu)"]},{"cell_type":"code","metadata":{"id":"mmvYJvgtiQz-","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":714},"outputId":"40003ea4-cbdc-4834-b2b5-3256a04e6835","executionInfo":{"status":"ok","timestamp":1577084823554,"user_tz":-60,"elapsed":45874,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='tanh', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='tanh'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer=optimizers.RMSprop(lr=0.001),\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=20,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":72,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/20\n","15000/15000 [==============================] - 5s 318us/step - loss: 0.4809 - acc: 0.7992 - val_loss: 0.3575 - val_acc: 0.8752\n","Epoch 2/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.2724 - acc: 0.9122 - val_loss: 0.2904 - val_acc: 0.8861\n","Epoch 3/20\n","15000/15000 [==============================] - 2s 142us/step - loss: 0.1948 - acc: 0.9338 - val_loss: 0.2703 - val_acc: 0.8880\n","Epoch 4/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.1463 - acc: 0.9526 - val_loss: 0.2918 - val_acc: 0.8852\n","Epoch 5/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.1122 - acc: 0.9636 - val_loss: 0.3039 - val_acc: 0.8821\n","Epoch 6/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0871 - acc: 0.9718 - val_loss: 0.3885 - val_acc: 0.8666\n","Epoch 7/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0693 - acc: 0.9787 - val_loss: 0.3738 - val_acc: 0.8746\n","Epoch 8/20\n","15000/15000 [==============================] - 2s 144us/step - loss: 0.0513 - acc: 0.9855 - val_loss: 0.4613 - val_acc: 0.8687\n","Epoch 9/20\n","15000/15000 [==============================] - 2s 142us/step - loss: 0.0464 - acc: 0.9851 - val_loss: 0.4683 - val_acc: 0.8744\n","Epoch 10/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0367 - acc: 0.9886 - val_loss: 0.4962 - val_acc: 0.8726\n","Epoch 11/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0284 - acc: 0.9929 - val_loss: 0.5416 - val_acc: 0.8686\n","Epoch 12/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0118 - acc: 0.9985 - val_loss: 0.5776 - val_acc: 0.8645\n","Epoch 13/20\n","15000/15000 [==============================] - 2s 140us/step - loss: 0.0225 - acc: 0.9939 - val_loss: 0.6091 - val_acc: 0.8670\n","Epoch 14/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0047 - acc: 0.9996 - val_loss: 0.6488 - val_acc: 0.8644\n","Epoch 15/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0181 - acc: 0.9948 - val_loss: 0.6757 - val_acc: 0.8667\n","Epoch 16/20\n","15000/15000 [==============================] - 2s 142us/step - loss: 0.0018 - acc: 0.9999 - val_loss: 0.7101 - val_acc: 0.8644\n","Epoch 17/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 0.0012 - acc: 1.0000 - val_loss: 0.9683 - val_acc: 0.8347\n","Epoch 18/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0138 - acc: 0.9959 - val_loss: 0.7746 - val_acc: 0.8640\n","Epoch 19/20\n","15000/15000 [==============================] - 2s 139us/step - loss: 6.0908e-04 - acc: 1.0000 - val_loss: 0.8091 - val_acc: 0.8627\n","Epoch 20/20\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.0032 - acc: 0.9992 - val_loss: 0.9530 - val_acc: 0.8466\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"YvLE-4MUk_L2","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":280},"outputId":"3209c3fa-90d7-4233-ef45-299e0a63cef7","executionInfo":{"status":"ok","timestamp":1577084842607,"user_tz":-60,"elapsed":938,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["history_dict= history.history\n","visualize_history(history_dict)"],"execution_count":73,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAwgAAAEHCAYAAAAKzGHrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n1gOTXAYAAAAASUVORK5CYII=\n","text/plain":["<Figure size 936x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"id":"-WG8FoOslO5g","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":136},"outputId":"ee8e56da-dffc-4e3c-d0c8-ee81f5cb2602","executionInfo":{"status":"ok","timestamp":1577084867985,"user_tz":-60,"elapsed":8461,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(16,activation='tanh', input_shape=(10000,)))\n","model.add(layers.Dense(16,activation='tanh'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer=optimizers.RMSprop(lr=0.001),\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=3,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":74,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/3\n","15000/15000 [==============================] - 3s 220us/step - loss: 0.4983 - acc: 0.7973 - val_loss: 0.3776 - val_acc: 0.8600\n","Epoch 2/3\n","15000/15000 [==============================] - 2s 142us/step - loss: 0.2846 - acc: 0.9074 - val_loss: 0.2874 - val_acc: 0.8892\n","Epoch 3/3\n","15000/15000 [==============================] - 2s 141us/step - loss: 0.2004 - acc: 0.9332 - val_loss: 0.2732 - val_acc: 0.8879\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"fuaiRu4YlTQl","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":51},"outputId":"45ffa7f3-031f-4dc6-d2fe-78229b55253e","executionInfo":{"status":"ok","timestamp":1577084875083,"user_tz":-60,"elapsed":4949,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["results = model.evaluate(x_test,y_test)\n","results"],"execution_count":75,"outputs":[{"output_type":"stream","text":["25000/25000 [==============================] - 5s 185us/step\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["[0.2964186774349213, 0.87884]"]},"metadata":{"tags":[]},"execution_count":75}]},{"cell_type":"markdown","metadata":{"id":"Gz212XeKlnLL","colab_type":"text"},"source":["## Neural network 6\n","\n","- 4 hidden layers (instead of 2)\n","- number of units: 64 (instead of 16)\n","- loss function: binary_crossentropy\n","- activation function: relu"]},{"cell_type":"code","metadata":{"id":"2QHbNdcGlV1Q","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":714},"outputId":"c9998b76-7c53-4160-ad7f-a72a459de2d0","executionInfo":{"status":"ok","timestamp":1577084981245,"user_tz":-60,"elapsed":61846,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(64,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer=optimizers.RMSprop(lr=0.001),\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=20,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":76,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/20\n","15000/15000 [==============================] - 4s 295us/step - loss: 0.5117 - acc: 0.7426 - val_loss: 0.4043 - val_acc: 0.8282\n","Epoch 2/20\n","15000/15000 [==============================] - 3s 198us/step - loss: 0.2853 - acc: 0.8890 - val_loss: 0.2839 - val_acc: 0.8844\n","Epoch 3/20\n","15000/15000 [==============================] - 3s 199us/step - loss: 0.2063 - acc: 0.9218 - val_loss: 0.2762 - val_acc: 0.8900\n","Epoch 4/20\n","15000/15000 [==============================] - 3s 197us/step - loss: 0.1541 - acc: 0.9418 - val_loss: 0.3123 - val_acc: 0.8841\n","Epoch 5/20\n","15000/15000 [==============================] - 3s 200us/step - loss: 0.1122 - acc: 0.9615 - val_loss: 0.4474 - val_acc: 0.8646\n","Epoch 6/20\n","15000/15000 [==============================] - 3s 198us/step - loss: 0.1073 - acc: 0.9582 - val_loss: 0.3692 - val_acc: 0.8839\n","Epoch 7/20\n","15000/15000 [==============================] - 3s 200us/step - loss: 0.0928 - acc: 0.9737 - val_loss: 0.3964 - val_acc: 0.8791\n","Epoch 8/20\n","15000/15000 [==============================] - 3s 200us/step - loss: 0.0163 - acc: 0.9967 - val_loss: 0.5092 - val_acc: 0.8754\n","Epoch 9/20\n","15000/15000 [==============================] - 3s 199us/step - loss: 0.0869 - acc: 0.9822 - val_loss: 0.5279 - val_acc: 0.8795\n","Epoch 10/20\n","15000/15000 [==============================] - 3s 199us/step - loss: 0.0037 - acc: 0.9997 - val_loss: 0.6419 - val_acc: 0.8785\n","Epoch 11/20\n","15000/15000 [==============================] - 3s 197us/step - loss: 0.1137 - acc: 0.9823 - val_loss: 0.6827 - val_acc: 0.8744\n","Epoch 12/20\n","15000/15000 [==============================] - 3s 200us/step - loss: 0.0010 - acc: 0.9999 - val_loss: 0.7278 - val_acc: 0.8748\n","Epoch 13/20\n","15000/15000 [==============================] - 3s 197us/step - loss: 5.6827e-04 - acc: 0.9999 - val_loss: 0.7988 - val_acc: 0.8742\n","Epoch 14/20\n","15000/15000 [==============================] - 3s 197us/step - loss: 2.5500e-04 - acc: 0.9999 - val_loss: 0.9093 - val_acc: 0.8736\n","Epoch 15/20\n","15000/15000 [==============================] - 3s 197us/step - loss: 5.7288e-05 - acc: 1.0000 - val_loss: 1.0354 - val_acc: 0.8735\n","Epoch 16/20\n","15000/15000 [==============================] - 3s 196us/step - loss: 1.4239e-05 - acc: 1.0000 - val_loss: 1.1209 - val_acc: 0.8727\n","Epoch 17/20\n","15000/15000 [==============================] - 3s 198us/step - loss: 3.6432e-06 - acc: 1.0000 - val_loss: 1.2000 - val_acc: 0.8715\n","Epoch 18/20\n","15000/15000 [==============================] - 3s 198us/step - loss: 7.4989e-07 - acc: 1.0000 - val_loss: 1.2691 - val_acc: 0.8717\n","Epoch 19/20\n","15000/15000 [==============================] - 3s 196us/step - loss: 2.0992e-07 - acc: 1.0000 - val_loss: 1.3118 - val_acc: 0.8713\n","Epoch 20/20\n","15000/15000 [==============================] - 3s 193us/step - loss: 1.3524e-07 - acc: 1.0000 - val_loss: 1.3374 - val_acc: 0.8714\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"QEspgwOflh4H","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":283},"outputId":"7f414877-f607-4ed5-bbb1-3fade9ed0612","executionInfo":{"status":"ok","timestamp":1577085003361,"user_tz":-60,"elapsed":820,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["history_dict= history.history\n","visualize_history(history_dict)"],"execution_count":77,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAwgAAAEKCAYAAAC2UnI1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nNZxLN4EAAAAASUVORK5CYII=\n","text/plain":["<Figure size 936x288 with 2 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"id":"8m4GSAotl2GP","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":102},"outputId":"dc5bb6f1-626c-4df9-8920-135d1752d219","executionInfo":{"status":"ok","timestamp":1577085026585,"user_tz":-60,"elapsed":8308,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["model = models.Sequential()\n","model.add(layers.Dense(64,activation='relu', input_shape=(10000,)))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(64,activation='relu'))\n","model.add(layers.Dense(1,activation='sigmoid'))\n","\n","model.compile(optimizer=optimizers.RMSprop(lr=0.001),\n","                       loss='binary_crossentropy',\n","                       metrics=['accuracy']) \n","\n","history = model.fit(partial_x_train, \n","                              partial_y_train,\n","                              epochs=2,\n","                              batch_size=512,\n","                              validation_data=(x_val,y_val))"],"execution_count":78,"outputs":[{"output_type":"stream","text":["Train on 15000 samples, validate on 10000 samples\n","Epoch 1/2\n","15000/15000 [==============================] - 4s 293us/step - loss: 0.5209 - acc: 0.7425 - val_loss: 0.4656 - val_acc: 0.7998\n","Epoch 2/2\n","15000/15000 [==============================] - 3s 195us/step - loss: 0.2898 - acc: 0.8863 - val_loss: 0.3042 - val_acc: 0.8781\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"jAftGlIBl6BF","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":51},"outputId":"b6f7efde-00a3-4cd7-f442-6d282bcdb52f","executionInfo":{"status":"ok","timestamp":1577085030089,"user_tz":-60,"elapsed":2757,"user":{"displayName":"Kotaro Sonoda","photoUrl":"","userId":"14975548384331586092"}}},"source":["results = model.evaluate(x_test,y_test)\n","results"],"execution_count":79,"outputs":[{"output_type":"stream","text":["25000/25000 [==============================] - 2s 94us/step\n"],"name":"stdout"},{"output_type":"execute_result","data":{"text/plain":["[0.32402512489318847, 0.86912]"]},"metadata":{"tags":[]},"execution_count":79}]},{"cell_type":"code","metadata":{"id":"XkvpRzetl8Om","colab_type":"code","colab":{}},"source":[""],"execution_count":0,"outputs":[]}]}