notebook.community

Edit and run 





In [1]:



    


from keras.models import Sequential
from keras.layers import Dense
import numpy as np



    


















    






Using TensorFlow backend.

















In [2]:



    


np.random.seed(7)



    











In [3]:



    


dataset = np.loadtxt('pima-indians-diabetes.data', delimiter=',')
X = dataset[:, 0:8]
Y = dataset[:, 8]



    











In [4]:



    


model = Sequential()
model.add(Dense(12, input_dim=8, activation='relu'))
model.add(Dense(8, activation='relu'))
model.add(Dense(1, activation='sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])



    











In [8]:



    


model.fit(X, Y, validation_split=0.33, epochs=150, batch_size=10)



    


















    






Train on 514 samples, validate on 254 samples
Epoch 1/150
514/514 [==============================] - 0s - loss: 6.0377 - acc: 0.3852 - val_loss: 4.3065 - val_acc: 0.5118
Epoch 2/150
514/514 [==============================] - 0s - loss: 2.4940 - acc: 0.5409 - val_loss: 1.9934 - val_acc: 0.5118
Epoch 3/150
514/514 [==============================] - 0s - loss: 1.5921 - acc: 0.5603 - val_loss: 1.7380 - val_acc: 0.5197
Epoch 4/150
514/514 [==============================] - 0s - loss: 1.3562 - acc: 0.5720 - val_loss: 1.4103 - val_acc: 0.5354
Epoch 5/150
514/514 [==============================] - 0s - loss: 1.1636 - acc: 0.5895 - val_loss: 1.3189 - val_acc: 0.5039
Epoch 6/150
514/514 [==============================] - 0s - loss: 1.0265 - acc: 0.5642 - val_loss: 1.1053 - val_acc: 0.5433
Epoch 7/150
514/514 [==============================] - 0s - loss: 0.9333 - acc: 0.5856 - val_loss: 1.0059 - val_acc: 0.5827
Epoch 8/150
514/514 [==============================] - 0s - loss: 0.8754 - acc: 0.6109 - val_loss: 0.9774 - val_acc: 0.5669
Epoch 9/150
514/514 [==============================] - 0s - loss: 0.8073 - acc: 0.6187 - val_loss: 0.9216 - val_acc: 0.6142
Epoch 10/150
514/514 [==============================] - 0s - loss: 0.7870 - acc: 0.6167 - val_loss: 0.8556 - val_acc: 0.6063
Epoch 11/150
514/514 [==============================] - 0s - loss: 0.7494 - acc: 0.6323 - val_loss: 0.8728 - val_acc: 0.6181
Epoch 12/150
514/514 [==============================] - 0s - loss: 0.7397 - acc: 0.6498 - val_loss: 0.8721 - val_acc: 0.6496
Epoch 13/150
514/514 [==============================] - 0s - loss: 0.6973 - acc: 0.6576 - val_loss: 0.8084 - val_acc: 0.6339
Epoch 14/150
514/514 [==============================] - 0s - loss: 0.7233 - acc: 0.6459 - val_loss: 0.8645 - val_acc: 0.6102
Epoch 15/150
514/514 [==============================] - 0s - loss: 0.6903 - acc: 0.6712 - val_loss: 0.7862 - val_acc: 0.6417
Epoch 16/150
514/514 [==============================] - 0s - loss: 0.6798 - acc: 0.6712 - val_loss: 0.7577 - val_acc: 0.6220
Epoch 17/150
514/514 [==============================] - 0s - loss: 0.6971 - acc: 0.6595 - val_loss: 0.7647 - val_acc: 0.6417
Epoch 18/150
514/514 [==============================] - 0s - loss: 0.6533 - acc: 0.6712 - val_loss: 0.7759 - val_acc: 0.6181
Epoch 19/150
514/514 [==============================] - 0s - loss: 0.6643 - acc: 0.6907 - val_loss: 0.7405 - val_acc: 0.6417
Epoch 20/150
514/514 [==============================] - 0s - loss: 0.6687 - acc: 0.6751 - val_loss: 0.7224 - val_acc: 0.6535
Epoch 21/150
514/514 [==============================] - 0s - loss: 0.6264 - acc: 0.6829 - val_loss: 0.7075 - val_acc: 0.6535
Epoch 22/150
514/514 [==============================] - 0s - loss: 0.6632 - acc: 0.6907 - val_loss: 0.7352 - val_acc: 0.6614
Epoch 23/150
514/514 [==============================] - 0s - loss: 0.6391 - acc: 0.6770 - val_loss: 0.7112 - val_acc: 0.6732
Epoch 24/150
514/514 [==============================] - 0s - loss: 0.6227 - acc: 0.6907 - val_loss: 0.7241 - val_acc: 0.6457
Epoch 25/150
514/514 [==============================] - 0s - loss: 0.6485 - acc: 0.6693 - val_loss: 0.6954 - val_acc: 0.6496
Epoch 26/150
514/514 [==============================] - 0s - loss: 0.6084 - acc: 0.7023 - val_loss: 0.7477 - val_acc: 0.6457
Epoch 27/150
514/514 [==============================] - 0s - loss: 0.6227 - acc: 0.6946 - val_loss: 0.7741 - val_acc: 0.6535
Epoch 28/150
514/514 [==============================] - 0s - loss: 0.6088 - acc: 0.7004 - val_loss: 0.6983 - val_acc: 0.6850
Epoch 29/150
514/514 [==============================] - 0s - loss: 0.6352 - acc: 0.6712 - val_loss: 0.6995 - val_acc: 0.6732
Epoch 30/150
514/514 [==============================] - 0s - loss: 0.6027 - acc: 0.6946 - val_loss: 0.6734 - val_acc: 0.6457
Epoch 31/150
514/514 [==============================] - 0s - loss: 0.6113 - acc: 0.6984 - val_loss: 0.7503 - val_acc: 0.6535
Epoch 32/150
514/514 [==============================] - 0s - loss: 0.5939 - acc: 0.6984 - val_loss: 0.6500 - val_acc: 0.6969
Epoch 33/150
514/514 [==============================] - 0s - loss: 0.5920 - acc: 0.6887 - val_loss: 0.7098 - val_acc: 0.6732
Epoch 34/150
514/514 [==============================] - 0s - loss: 0.6464 - acc: 0.6790 - val_loss: 0.6538 - val_acc: 0.6850
Epoch 35/150
514/514 [==============================] - 0s - loss: 0.5878 - acc: 0.7276 - val_loss: 0.6414 - val_acc: 0.6929
Epoch 36/150
514/514 [==============================] - 0s - loss: 0.5998 - acc: 0.7160 - val_loss: 0.6339 - val_acc: 0.6929
Epoch 37/150
514/514 [==============================] - 0s - loss: 0.5907 - acc: 0.7023 - val_loss: 0.6194 - val_acc: 0.7087
Epoch 38/150
514/514 [==============================] - 0s - loss: 0.5708 - acc: 0.7276 - val_loss: 0.6212 - val_acc: 0.7047
Epoch 39/150
514/514 [==============================] - 0s - loss: 0.6052 - acc: 0.6984 - val_loss: 0.6567 - val_acc: 0.6969
Epoch 40/150
514/514 [==============================] - 0s - loss: 0.6039 - acc: 0.6965 - val_loss: 0.6598 - val_acc: 0.6732
Epoch 41/150
514/514 [==============================] - 0s - loss: 0.5963 - acc: 0.6984 - val_loss: 0.6576 - val_acc: 0.6850
Epoch 42/150
514/514 [==============================] - 0s - loss: 0.6217 - acc: 0.7257 - val_loss: 0.6581 - val_acc: 0.6850
Epoch 43/150
514/514 [==============================] - 0s - loss: 0.5799 - acc: 0.7062 - val_loss: 0.6287 - val_acc: 0.7008
Epoch 44/150
514/514 [==============================] - 0s - loss: 0.5630 - acc: 0.7198 - val_loss: 0.6621 - val_acc: 0.6693
Epoch 45/150
514/514 [==============================] - 0s - loss: 0.5967 - acc: 0.7062 - val_loss: 0.6510 - val_acc: 0.6929
Epoch 46/150
514/514 [==============================] - 0s - loss: 0.6639 - acc: 0.7062 - val_loss: 0.6187 - val_acc: 0.7165
Epoch 47/150
514/514 [==============================] - 0s - loss: 0.5891 - acc: 0.7004 - val_loss: 0.6079 - val_acc: 0.7362
Epoch 48/150
514/514 [==============================] - 0s - loss: 0.5949 - acc: 0.7121 - val_loss: 0.6315 - val_acc: 0.7087
Epoch 49/150
514/514 [==============================] - 0s - loss: 0.5747 - acc: 0.7179 - val_loss: 0.6226 - val_acc: 0.6929
Epoch 50/150
514/514 [==============================] - 0s - loss: 0.5634 - acc: 0.7179 - val_loss: 0.6289 - val_acc: 0.7008
Epoch 51/150
514/514 [==============================] - 0s - loss: 0.5804 - acc: 0.7101 - val_loss: 0.6843 - val_acc: 0.6850
Epoch 52/150
514/514 [==============================] - 0s - loss: 0.5792 - acc: 0.7101 - val_loss: 0.6947 - val_acc: 0.6693
Epoch 53/150
514/514 [==============================] - 0s - loss: 0.6082 - acc: 0.7062 - val_loss: 0.6342 - val_acc: 0.6969
Epoch 54/150
514/514 [==============================] - 0s - loss: 0.5952 - acc: 0.7160 - val_loss: 0.6449 - val_acc: 0.7047
Epoch 55/150
514/514 [==============================] - 0s - loss: 0.5592 - acc: 0.7335 - val_loss: 0.6223 - val_acc: 0.7362
Epoch 56/150
514/514 [==============================] - 0s - loss: 0.5631 - acc: 0.7374 - val_loss: 0.6191 - val_acc: 0.7008
Epoch 57/150
514/514 [==============================] - 0s - loss: 0.5892 - acc: 0.6946 - val_loss: 0.6368 - val_acc: 0.7441
Epoch 58/150
514/514 [==============================] - 0s - loss: 0.6010 - acc: 0.7121 - val_loss: 0.5870 - val_acc: 0.7441
Epoch 59/150
514/514 [==============================] - 0s - loss: 0.5910 - acc: 0.6946 - val_loss: 0.6003 - val_acc: 0.7047
Epoch 60/150
514/514 [==============================] - 0s - loss: 0.5902 - acc: 0.6965 - val_loss: 0.5918 - val_acc: 0.7283
Epoch 61/150
514/514 [==============================] - 0s - loss: 0.5518 - acc: 0.7451 - val_loss: 0.6086 - val_acc: 0.7165
Epoch 62/150
514/514 [==============================] - 0s - loss: 0.5456 - acc: 0.7432 - val_loss: 0.6439 - val_acc: 0.6850
Epoch 63/150
514/514 [==============================] - 0s - loss: 0.5788 - acc: 0.7257 - val_loss: 0.6009 - val_acc: 0.7402
Epoch 64/150
514/514 [==============================] - 0s - loss: 0.5712 - acc: 0.7315 - val_loss: 0.6342 - val_acc: 0.7008
Epoch 65/150
514/514 [==============================] - 0s - loss: 0.5477 - acc: 0.7393 - val_loss: 0.6095 - val_acc: 0.7244
Epoch 66/150
514/514 [==============================] - 0s - loss: 0.5357 - acc: 0.7529 - val_loss: 0.6311 - val_acc: 0.7165
Epoch 67/150
514/514 [==============================] - 0s - loss: 0.5707 - acc: 0.7179 - val_loss: 0.6251 - val_acc: 0.7126
Epoch 68/150
514/514 [==============================] - 0s - loss: 0.5493 - acc: 0.7198 - val_loss: 0.6012 - val_acc: 0.7165
Epoch 69/150
514/514 [==============================] - 0s - loss: 0.5646 - acc: 0.7140 - val_loss: 0.5979 - val_acc: 0.7165
Epoch 70/150
514/514 [==============================] - 0s - loss: 0.5684 - acc: 0.7140 - val_loss: 0.6257 - val_acc: 0.6929
Epoch 71/150
514/514 [==============================] - 0s - loss: 0.5675 - acc: 0.7354 - val_loss: 0.6413 - val_acc: 0.6850
Epoch 72/150
514/514 [==============================] - 0s - loss: 0.5641 - acc: 0.7198 - val_loss: 0.6093 - val_acc: 0.6969
Epoch 73/150
514/514 [==============================] - 0s - loss: 0.5663 - acc: 0.7374 - val_loss: 0.5954 - val_acc: 0.7126
Epoch 74/150
514/514 [==============================] - 0s - loss: 0.5756 - acc: 0.7510 - val_loss: 0.6067 - val_acc: 0.7402
Epoch 75/150
514/514 [==============================] - 0s - loss: 0.5591 - acc: 0.7179 - val_loss: 0.5778 - val_acc: 0.7283
Epoch 76/150
514/514 [==============================] - 0s - loss: 0.5493 - acc: 0.7568 - val_loss: 0.5987 - val_acc: 0.7244
Epoch 77/150
514/514 [==============================] - 0s - loss: 0.5427 - acc: 0.7315 - val_loss: 0.6136 - val_acc: 0.7087
Epoch 78/150
514/514 [==============================] - 0s - loss: 0.5495 - acc: 0.7471 - val_loss: 0.6016 - val_acc: 0.7165
Epoch 79/150
514/514 [==============================] - 0s - loss: 0.5748 - acc: 0.7237 - val_loss: 0.5972 - val_acc: 0.7087
Epoch 80/150
514/514 [==============================] - 0s - loss: 0.5570 - acc: 0.7121 - val_loss: 0.5873 - val_acc: 0.7520
Epoch 81/150
514/514 [==============================] - 0s - loss: 0.5633 - acc: 0.7335 - val_loss: 0.6893 - val_acc: 0.6811
Epoch 82/150
514/514 [==============================] - 0s - loss: 0.5575 - acc: 0.7451 - val_loss: 0.6777 - val_acc: 0.6417
Epoch 83/150
514/514 [==============================] - 0s - loss: 0.5510 - acc: 0.7529 - val_loss: 0.6004 - val_acc: 0.7205
Epoch 84/150
514/514 [==============================] - 0s - loss: 0.5339 - acc: 0.7549 - val_loss: 0.6571 - val_acc: 0.6929
Epoch 85/150
514/514 [==============================] - 0s - loss: 0.5481 - acc: 0.7374 - val_loss: 0.6087 - val_acc: 0.7047
Epoch 86/150
514/514 [==============================] - 0s - loss: 0.5507 - acc: 0.7335 - val_loss: 0.5764 - val_acc: 0.7441
Epoch 87/150
514/514 [==============================] - 0s - loss: 0.5370 - acc: 0.7315 - val_loss: 0.5848 - val_acc: 0.7205
Epoch 88/150
514/514 [==============================] - 0s - loss: 0.5514 - acc: 0.7276 - val_loss: 0.6322 - val_acc: 0.7126
Epoch 89/150
514/514 [==============================] - 0s - loss: 0.5583 - acc: 0.7374 - val_loss: 0.6930 - val_acc: 0.6929
Epoch 90/150
514/514 [==============================] - 0s - loss: 0.5517 - acc: 0.7432 - val_loss: 0.6209 - val_acc: 0.6929
Epoch 91/150
514/514 [==============================] - 0s - loss: 0.5502 - acc: 0.7276 - val_loss: 0.5909 - val_acc: 0.7283
Epoch 92/150
514/514 [==============================] - 0s - loss: 0.5587 - acc: 0.7490 - val_loss: 0.6103 - val_acc: 0.7165
Epoch 93/150
514/514 [==============================] - 0s - loss: 0.5625 - acc: 0.7257 - val_loss: 0.7228 - val_acc: 0.6850
Epoch 94/150
514/514 [==============================] - 0s - loss: 0.5532 - acc: 0.7296 - val_loss: 0.6420 - val_acc: 0.6732
Epoch 95/150
514/514 [==============================] - 0s - loss: 0.6471 - acc: 0.7179 - val_loss: 0.7324 - val_acc: 0.6811
Epoch 96/150
514/514 [==============================] - 0s - loss: 0.5711 - acc: 0.7471 - val_loss: 0.5692 - val_acc: 0.7205
Epoch 97/150
514/514 [==============================] - 0s - loss: 0.6085 - acc: 0.7198 - val_loss: 0.6056 - val_acc: 0.6969
Epoch 98/150
514/514 [==============================] - 0s - loss: 0.5433 - acc: 0.7393 - val_loss: 0.5737 - val_acc: 0.7205
Epoch 99/150
514/514 [==============================] - 0s - loss: 0.5472 - acc: 0.7432 - val_loss: 0.5880 - val_acc: 0.7441
Epoch 100/150
514/514 [==============================] - 0s - loss: 0.5965 - acc: 0.7374 - val_loss: 0.6206 - val_acc: 0.7008
Epoch 101/150
514/514 [==============================] - 0s - loss: 0.5803 - acc: 0.7237 - val_loss: 0.5956 - val_acc: 0.7047
Epoch 102/150
514/514 [==============================] - 0s - loss: 0.5510 - acc: 0.7374 - val_loss: 0.6116 - val_acc: 0.7087
Epoch 103/150
514/514 [==============================] - 0s - loss: 0.5517 - acc: 0.7121 - val_loss: 0.5846 - val_acc: 0.7244
Epoch 104/150
514/514 [==============================] - 0s - loss: 0.5510 - acc: 0.7354 - val_loss: 0.5874 - val_acc: 0.7283
Epoch 105/150
514/514 [==============================] - 0s - loss: 0.5433 - acc: 0.7276 - val_loss: 0.6032 - val_acc: 0.6969
Epoch 106/150
514/514 [==============================] - 0s - loss: 0.5386 - acc: 0.7451 - val_loss: 0.5885 - val_acc: 0.7205
Epoch 107/150
514/514 [==============================] - 0s - loss: 0.5253 - acc: 0.7451 - val_loss: 0.6288 - val_acc: 0.7047
Epoch 108/150
514/514 [==============================] - 0s - loss: 0.5470 - acc: 0.7237 - val_loss: 0.5788 - val_acc: 0.7283
Epoch 109/150
514/514 [==============================] - 0s - loss: 0.5319 - acc: 0.7529 - val_loss: 0.6064 - val_acc: 0.7244
Epoch 110/150
514/514 [==============================] - 0s - loss: 0.5320 - acc: 0.7568 - val_loss: 0.5848 - val_acc: 0.7283
Epoch 111/150
514/514 [==============================] - 0s - loss: 0.5412 - acc: 0.7354 - val_loss: 0.7940 - val_acc: 0.6457
Epoch 112/150
514/514 [==============================] - 0s - loss: 0.5516 - acc: 0.7043 - val_loss: 0.5837 - val_acc: 0.7165
Epoch 113/150
514/514 [==============================] - 0s - loss: 0.5393 - acc: 0.7257 - val_loss: 0.5716 - val_acc: 0.7205
Epoch 114/150
514/514 [==============================] - 0s - loss: 0.5412 - acc: 0.7296 - val_loss: 0.5985 - val_acc: 0.7087
Epoch 115/150
514/514 [==============================] - 0s - loss: 0.5306 - acc: 0.7607 - val_loss: 0.5860 - val_acc: 0.7087
Epoch 116/150
514/514 [==============================] - 0s - loss: 0.5840 - acc: 0.7335 - val_loss: 0.6217 - val_acc: 0.7205
Epoch 117/150
514/514 [==============================] - 0s - loss: 0.5462 - acc: 0.7549 - val_loss: 0.5881 - val_acc: 0.7087
Epoch 118/150
514/514 [==============================] - 0s - loss: 0.5441 - acc: 0.7140 - val_loss: 0.6008 - val_acc: 0.7126
Epoch 119/150
514/514 [==============================] - 0s - loss: 0.5695 - acc: 0.7432 - val_loss: 0.6690 - val_acc: 0.6890
Epoch 120/150
514/514 [==============================] - 0s - loss: 0.5343 - acc: 0.7568 - val_loss: 0.5713 - val_acc: 0.7244
Epoch 121/150
514/514 [==============================] - 0s - loss: 0.5569 - acc: 0.7179 - val_loss: 0.5848 - val_acc: 0.7244
Epoch 122/150
514/514 [==============================] - 0s - loss: 0.5825 - acc: 0.7257 - val_loss: 0.6380 - val_acc: 0.6969
Epoch 123/150
514/514 [==============================] - 0s - loss: 0.5548 - acc: 0.7451 - val_loss: 0.5774 - val_acc: 0.7283
Epoch 124/150
514/514 [==============================] - 0s - loss: 0.5418 - acc: 0.7374 - val_loss: 0.5629 - val_acc: 0.7402
Epoch 125/150
514/514 [==============================] - 0s - loss: 0.5259 - acc: 0.7432 - val_loss: 0.5900 - val_acc: 0.7047
Epoch 126/150
514/514 [==============================] - 0s - loss: 0.5528 - acc: 0.7257 - val_loss: 0.6274 - val_acc: 0.7205
Epoch 127/150
514/514 [==============================] - 0s - loss: 0.5297 - acc: 0.7412 - val_loss: 0.6106 - val_acc: 0.7441
Epoch 128/150
514/514 [==============================] - 0s - loss: 0.5692 - acc: 0.7218 - val_loss: 0.7843 - val_acc: 0.6417
Epoch 129/150
514/514 [==============================] - 0s - loss: 0.5344 - acc: 0.7412 - val_loss: 0.5721 - val_acc: 0.7205
Epoch 130/150
514/514 [==============================] - 0s - loss: 0.5273 - acc: 0.7568 - val_loss: 0.5664 - val_acc: 0.7126
Epoch 131/150
514/514 [==============================] - 0s - loss: 0.5289 - acc: 0.7296 - val_loss: 0.5726 - val_acc: 0.7087
Epoch 132/150
514/514 [==============================] - 0s - loss: 0.5304 - acc: 0.7490 - val_loss: 0.5856 - val_acc: 0.7323
Epoch 133/150
514/514 [==============================] - 0s - loss: 0.5126 - acc: 0.7451 - val_loss: 0.5728 - val_acc: 0.7087
Epoch 134/150
514/514 [==============================] - 0s - loss: 0.5151 - acc: 0.7607 - val_loss: 0.5663 - val_acc: 0.7283
Epoch 135/150
514/514 [==============================] - 0s - loss: 0.5099 - acc: 0.7588 - val_loss: 0.5630 - val_acc: 0.7283
Epoch 136/150
514/514 [==============================] - 0s - loss: 0.5170 - acc: 0.7432 - val_loss: 0.5635 - val_acc: 0.7165
Epoch 137/150
514/514 [==============================] - 0s - loss: 0.5151 - acc: 0.7374 - val_loss: 0.5601 - val_acc: 0.7205
Epoch 138/150
514/514 [==============================] - 0s - loss: 0.5089 - acc: 0.7529 - val_loss: 0.5590 - val_acc: 0.7283
Epoch 139/150
514/514 [==============================] - 0s - loss: 0.5137 - acc: 0.7451 - val_loss: 0.5604 - val_acc: 0.7165
Epoch 140/150
514/514 [==============================] - 0s - loss: 0.5300 - acc: 0.7257 - val_loss: 0.5730 - val_acc: 0.7402
Epoch 141/150
514/514 [==============================] - 0s - loss: 0.5072 - acc: 0.7646 - val_loss: 0.5581 - val_acc: 0.7598
Epoch 142/150
514/514 [==============================] - 0s - loss: 0.5404 - acc: 0.7607 - val_loss: 0.6411 - val_acc: 0.7008
Epoch 143/150
514/514 [==============================] - 0s - loss: 0.5328 - acc: 0.7296 - val_loss: 0.5693 - val_acc: 0.7402
Epoch 144/150
514/514 [==============================] - 0s - loss: 0.5143 - acc: 0.7626 - val_loss: 0.5983 - val_acc: 0.7362
Epoch 145/150
514/514 [==============================] - 0s - loss: 0.5388 - acc: 0.7451 - val_loss: 0.5664 - val_acc: 0.7283
Epoch 146/150
514/514 [==============================] - 0s - loss: 0.5115 - acc: 0.7374 - val_loss: 0.5625 - val_acc: 0.7283
Epoch 147/150
514/514 [==============================] - 0s - loss: 0.5020 - acc: 0.7549 - val_loss: 0.5858 - val_acc: 0.7126
Epoch 148/150
514/514 [==============================] - 0s - loss: 0.5241 - acc: 0.7374 - val_loss: 0.6635 - val_acc: 0.6890
Epoch 149/150
514/514 [==============================] - 0s - loss: 0.5241 - acc: 0.7490 - val_loss: 0.5511 - val_acc: 0.7480
Epoch 150/150
514/514 [==============================] - 0s - loss: 0.5221 - acc: 0.7607 - val_loss: 0.5785 - val_acc: 0.7087










    
Out[8]:








<keras.callbacks.History at 0x11fb36e80>

















In [5]:



    


from sklearn.model_selection import train_test_split



    











In [7]:



    


seed = 7
dataset = np.loadtxt('pima-indians-diabetes.data', delimiter=',')
X = dataset[:, 0:8]
Y = dataset[:, 8]
X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.33, random_state=seed)



    











In [10]:



    


model = Sequential()
model.add(Dense(12, input_dim=8, activation='relu'))
model.add(Dense(8, activation='relu'))
model.add(Dense(1, activation='sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])



    











In [12]:



    


model.fit(X_train, y_train, validation_data=(X_test, y_test), epochs=150, batch_size=10)



    


















    






Train on 514 samples, validate on 254 samples
Epoch 1/150
514/514 [==============================] - 0s - loss: 5.4229 - acc: 0.6304 - val_loss: 5.6229 - val_acc: 0.6378
Epoch 2/150
514/514 [==============================] - 0s - loss: 5.3085 - acc: 0.6401 - val_loss: 5.5007 - val_acc: 0.6417
Epoch 3/150
514/514 [==============================] - 0s - loss: 5.2346 - acc: 0.6459 - val_loss: 5.4103 - val_acc: 0.6378
Epoch 4/150
514/514 [==============================] - 0s - loss: 5.1298 - acc: 0.6323 - val_loss: 5.3088 - val_acc: 0.6378
Epoch 5/150
514/514 [==============================] - 0s - loss: 5.0421 - acc: 0.6401 - val_loss: 5.2040 - val_acc: 0.6378
Epoch 6/150
514/514 [==============================] - 0s - loss: 4.9210 - acc: 0.6323 - val_loss: 4.8707 - val_acc: 0.6378
Epoch 7/150
514/514 [==============================] - 0s - loss: 4.5375 - acc: 0.6498 - val_loss: 4.4086 - val_acc: 0.6220
Epoch 8/150
514/514 [==============================] - 0s - loss: 4.0227 - acc: 0.6128 - val_loss: 3.7692 - val_acc: 0.6063
Epoch 9/150
514/514 [==============================] - 0s - loss: 3.6753 - acc: 0.5992 - val_loss: 3.4825 - val_acc: 0.6260
Epoch 10/150
514/514 [==============================] - 0s - loss: 3.4646 - acc: 0.6420 - val_loss: 3.5302 - val_acc: 0.6378
Epoch 11/150
514/514 [==============================] - 0s - loss: 3.4321 - acc: 0.6245 - val_loss: 3.2797 - val_acc: 0.6496
Epoch 12/150
514/514 [==============================] - 0s - loss: 3.1151 - acc: 0.6342 - val_loss: 2.3772 - val_acc: 0.6378
Epoch 13/150
514/514 [==============================] - 0s - loss: 1.6956 - acc: 0.5486 - val_loss: 1.0138 - val_acc: 0.5866
Epoch 14/150
514/514 [==============================] - 0s - loss: 1.0465 - acc: 0.6051 - val_loss: 0.8784 - val_acc: 0.6220
Epoch 15/150
514/514 [==============================] - 0s - loss: 0.9950 - acc: 0.5953 - val_loss: 0.8158 - val_acc: 0.6417
Epoch 16/150
514/514 [==============================] - 0s - loss: 0.8745 - acc: 0.6498 - val_loss: 0.8259 - val_acc: 0.6339
Epoch 17/150
514/514 [==============================] - 0s - loss: 0.8333 - acc: 0.6770 - val_loss: 0.7215 - val_acc: 0.6890
Epoch 18/150
514/514 [==============================] - 0s - loss: 0.7670 - acc: 0.6673 - val_loss: 0.8371 - val_acc: 0.6772
Epoch 19/150
514/514 [==============================] - 0s - loss: 0.7134 - acc: 0.6887 - val_loss: 0.7544 - val_acc: 0.6693
Epoch 20/150
514/514 [==============================] - 0s - loss: 0.6996 - acc: 0.6770 - val_loss: 0.6842 - val_acc: 0.6693
Epoch 21/150
514/514 [==============================] - 0s - loss: 0.6600 - acc: 0.7043 - val_loss: 0.7152 - val_acc: 0.6772
Epoch 22/150
514/514 [==============================] - 0s - loss: 0.6240 - acc: 0.6907 - val_loss: 0.8918 - val_acc: 0.6535
Epoch 23/150
514/514 [==============================] - 0s - loss: 0.6177 - acc: 0.7121 - val_loss: 0.6894 - val_acc: 0.6693
Epoch 24/150
514/514 [==============================] - 0s - loss: 0.6216 - acc: 0.7043 - val_loss: 0.7710 - val_acc: 0.6535
Epoch 25/150
514/514 [==============================] - 0s - loss: 0.6004 - acc: 0.7179 - val_loss: 0.7013 - val_acc: 0.6654
Epoch 26/150
514/514 [==============================] - 0s - loss: 0.6341 - acc: 0.7004 - val_loss: 0.7162 - val_acc: 0.6575
Epoch 27/150
514/514 [==============================] - 0s - loss: 0.6112 - acc: 0.7062 - val_loss: 0.6918 - val_acc: 0.6693
Epoch 28/150
514/514 [==============================] - 0s - loss: 0.6237 - acc: 0.7004 - val_loss: 0.6774 - val_acc: 0.6811
Epoch 29/150
514/514 [==============================] - 0s - loss: 0.5800 - acc: 0.7140 - val_loss: 0.7153 - val_acc: 0.6614
Epoch 30/150
514/514 [==============================] - 0s - loss: 0.5761 - acc: 0.7335 - val_loss: 0.6456 - val_acc: 0.6929
Epoch 31/150
514/514 [==============================] - 0s - loss: 0.6061 - acc: 0.6965 - val_loss: 0.7567 - val_acc: 0.6457
Epoch 32/150
514/514 [==============================] - 0s - loss: 0.6030 - acc: 0.7082 - val_loss: 0.6793 - val_acc: 0.6654
Epoch 33/150
514/514 [==============================] - 0s - loss: 0.5941 - acc: 0.7354 - val_loss: 0.7651 - val_acc: 0.6575
Epoch 34/150
514/514 [==============================] - 0s - loss: 0.5608 - acc: 0.7257 - val_loss: 0.6485 - val_acc: 0.6969
Epoch 35/150
514/514 [==============================] - 0s - loss: 0.5502 - acc: 0.7374 - val_loss: 0.7316 - val_acc: 0.6575
Epoch 36/150
514/514 [==============================] - 0s - loss: 0.5767 - acc: 0.7510 - val_loss: 0.6455 - val_acc: 0.6890
Epoch 37/150
514/514 [==============================] - 0s - loss: 0.5594 - acc: 0.7160 - val_loss: 0.6402 - val_acc: 0.6929
Epoch 38/150
514/514 [==============================] - 0s - loss: 0.5439 - acc: 0.7276 - val_loss: 0.6381 - val_acc: 0.7047
Epoch 39/150
514/514 [==============================] - 0s - loss: 0.5635 - acc: 0.7218 - val_loss: 0.7396 - val_acc: 0.6732
Epoch 40/150
514/514 [==============================] - 0s - loss: 0.5515 - acc: 0.7393 - val_loss: 0.7263 - val_acc: 0.6614
Epoch 41/150
514/514 [==============================] - 0s - loss: 0.5921 - acc: 0.7082 - val_loss: 0.6538 - val_acc: 0.6850
Epoch 42/150
514/514 [==============================] - 0s - loss: 0.5501 - acc: 0.7315 - val_loss: 0.6771 - val_acc: 0.6772
Epoch 43/150
514/514 [==============================] - 0s - loss: 0.5532 - acc: 0.7529 - val_loss: 0.6349 - val_acc: 0.7165
Epoch 44/150
514/514 [==============================] - 0s - loss: 0.5652 - acc: 0.7374 - val_loss: 0.6378 - val_acc: 0.7283
Epoch 45/150
514/514 [==============================] - 0s - loss: 0.5577 - acc: 0.7257 - val_loss: 0.6409 - val_acc: 0.6969
Epoch 46/150
514/514 [==============================] - 0s - loss: 0.5257 - acc: 0.7626 - val_loss: 0.7328 - val_acc: 0.6772
Epoch 47/150
514/514 [==============================] - 0s - loss: 0.5445 - acc: 0.7374 - val_loss: 0.6554 - val_acc: 0.6890
Epoch 48/150
514/514 [==============================] - 0s - loss: 0.5227 - acc: 0.7335 - val_loss: 0.6995 - val_acc: 0.6693
Epoch 49/150
514/514 [==============================] - 0s - loss: 0.5438 - acc: 0.7451 - val_loss: 0.6356 - val_acc: 0.7008
Epoch 50/150
514/514 [==============================] - 0s - loss: 0.5326 - acc: 0.7393 - val_loss: 0.6192 - val_acc: 0.7362
Epoch 51/150
514/514 [==============================] - 0s - loss: 0.5323 - acc: 0.7374 - val_loss: 0.6228 - val_acc: 0.7244
Epoch 52/150
514/514 [==============================] - 0s - loss: 0.5473 - acc: 0.7335 - val_loss: 0.6297 - val_acc: 0.7008
Epoch 53/150
514/514 [==============================] - 0s - loss: 0.5417 - acc: 0.7393 - val_loss: 0.7364 - val_acc: 0.6772
Epoch 54/150
514/514 [==============================] - 0s - loss: 0.5593 - acc: 0.7354 - val_loss: 0.6551 - val_acc: 0.7087
Epoch 55/150
514/514 [==============================] - 0s - loss: 0.5665 - acc: 0.7101 - val_loss: 0.7917 - val_acc: 0.6732
Epoch 56/150
514/514 [==============================] - 0s - loss: 0.5598 - acc: 0.7276 - val_loss: 0.7145 - val_acc: 0.6732
Epoch 57/150
514/514 [==============================] - 0s - loss: 0.5683 - acc: 0.7121 - val_loss: 0.6474 - val_acc: 0.6929
Epoch 58/150
514/514 [==============================] - 0s - loss: 0.5130 - acc: 0.7490 - val_loss: 0.6193 - val_acc: 0.7205
Epoch 59/150
514/514 [==============================] - 0s - loss: 0.5131 - acc: 0.7549 - val_loss: 0.6713 - val_acc: 0.6850
Epoch 60/150
514/514 [==============================] - 0s - loss: 0.5356 - acc: 0.7412 - val_loss: 0.6392 - val_acc: 0.6654
Epoch 61/150
514/514 [==============================] - 0s - loss: 0.5237 - acc: 0.7510 - val_loss: 0.6218 - val_acc: 0.7205
Epoch 62/150
514/514 [==============================] - 0s - loss: 0.5149 - acc: 0.7529 - val_loss: 0.6698 - val_acc: 0.6890
Epoch 63/150
514/514 [==============================] - 0s - loss: 0.5303 - acc: 0.7490 - val_loss: 0.6416 - val_acc: 0.7087
Epoch 64/150
514/514 [==============================] - 0s - loss: 0.5160 - acc: 0.7510 - val_loss: 0.6206 - val_acc: 0.7126
Epoch 65/150
514/514 [==============================] - 0s - loss: 0.5177 - acc: 0.7354 - val_loss: 0.6191 - val_acc: 0.7244
Epoch 66/150
514/514 [==============================] - 0s - loss: 0.5431 - acc: 0.7412 - val_loss: 0.6633 - val_acc: 0.6614
Epoch 67/150
514/514 [==============================] - 0s - loss: 0.5224 - acc: 0.7451 - val_loss: 0.6463 - val_acc: 0.6890
Epoch 68/150
514/514 [==============================] - 0s - loss: 0.5173 - acc: 0.7607 - val_loss: 0.6123 - val_acc: 0.7323
Epoch 69/150
514/514 [==============================] - 0s - loss: 0.5175 - acc: 0.7529 - val_loss: 0.6173 - val_acc: 0.7047
Epoch 70/150
514/514 [==============================] - 0s - loss: 0.5393 - acc: 0.7529 - val_loss: 0.6327 - val_acc: 0.6890
Epoch 71/150
514/514 [==============================] - 0s - loss: 0.5392 - acc: 0.7588 - val_loss: 0.6250 - val_acc: 0.7283
Epoch 72/150
514/514 [==============================] - 0s - loss: 0.5603 - acc: 0.7393 - val_loss: 0.6474 - val_acc: 0.6969
Epoch 73/150
514/514 [==============================] - 0s - loss: 0.5235 - acc: 0.7490 - val_loss: 0.6080 - val_acc: 0.7244
Epoch 74/150
514/514 [==============================] - 0s - loss: 0.5251 - acc: 0.7549 - val_loss: 0.6376 - val_acc: 0.6969
Epoch 75/150
514/514 [==============================] - 0s - loss: 0.5034 - acc: 0.7568 - val_loss: 0.6201 - val_acc: 0.7323
Epoch 76/150
514/514 [==============================] - 0s - loss: 0.5240 - acc: 0.7412 - val_loss: 0.6070 - val_acc: 0.7402
Epoch 77/150
514/514 [==============================] - 0s - loss: 0.5021 - acc: 0.7665 - val_loss: 0.6797 - val_acc: 0.6732
Epoch 78/150
514/514 [==============================] - 0s - loss: 0.4974 - acc: 0.7782 - val_loss: 0.6169 - val_acc: 0.7283
Epoch 79/150
514/514 [==============================] - 0s - loss: 0.5128 - acc: 0.7490 - val_loss: 0.6092 - val_acc: 0.7362
Epoch 80/150
514/514 [==============================] - 0s - loss: 0.5172 - acc: 0.7490 - val_loss: 0.6262 - val_acc: 0.7047
Epoch 81/150
514/514 [==============================] - 0s - loss: 0.4978 - acc: 0.7471 - val_loss: 0.6047 - val_acc: 0.7362
Epoch 82/150
514/514 [==============================] - 0s - loss: 0.5027 - acc: 0.7704 - val_loss: 0.6434 - val_acc: 0.6772
Epoch 83/150
514/514 [==============================] - 0s - loss: 0.5026 - acc: 0.7529 - val_loss: 0.6077 - val_acc: 0.7244
Epoch 84/150
514/514 [==============================] - 0s - loss: 0.5383 - acc: 0.7529 - val_loss: 0.6169 - val_acc: 0.7205
Epoch 85/150
514/514 [==============================] - 0s - loss: 0.4882 - acc: 0.7685 - val_loss: 0.6228 - val_acc: 0.7165
Epoch 86/150
514/514 [==============================] - 0s - loss: 0.4995 - acc: 0.7529 - val_loss: 0.6546 - val_acc: 0.6811
Epoch 87/150
514/514 [==============================] - 0s - loss: 0.5166 - acc: 0.7646 - val_loss: 0.6051 - val_acc: 0.7283
Epoch 88/150
514/514 [==============================] - 0s - loss: 0.5017 - acc: 0.7510 - val_loss: 0.6108 - val_acc: 0.7126
Epoch 89/150
514/514 [==============================] - 0s - loss: 0.5304 - acc: 0.7626 - val_loss: 0.6625 - val_acc: 0.6850
Epoch 90/150
514/514 [==============================] - 0s - loss: 0.5193 - acc: 0.7490 - val_loss: 0.6013 - val_acc: 0.7283
Epoch 91/150
514/514 [==============================] - 0s - loss: 0.4943 - acc: 0.7743 - val_loss: 0.6079 - val_acc: 0.7283
Epoch 92/150
514/514 [==============================] - 0s - loss: 0.5184 - acc: 0.7529 - val_loss: 0.6213 - val_acc: 0.7126
Epoch 93/150
514/514 [==============================] - 0s - loss: 0.5042 - acc: 0.7451 - val_loss: 0.6311 - val_acc: 0.6811
Epoch 94/150
514/514 [==============================] - 0s - loss: 0.4954 - acc: 0.7607 - val_loss: 0.6025 - val_acc: 0.7323
Epoch 95/150
514/514 [==============================] - 0s - loss: 0.4917 - acc: 0.7529 - val_loss: 0.6167 - val_acc: 0.7205
Epoch 96/150
514/514 [==============================] - 0s - loss: 0.4894 - acc: 0.7490 - val_loss: 0.6030 - val_acc: 0.7441
Epoch 97/150
514/514 [==============================] - 0s - loss: 0.5054 - acc: 0.7471 - val_loss: 0.6028 - val_acc: 0.7283
Epoch 98/150
514/514 [==============================] - 0s - loss: 0.4863 - acc: 0.7840 - val_loss: 0.6134 - val_acc: 0.7244
Epoch 99/150
514/514 [==============================] - 0s - loss: 0.4870 - acc: 0.7471 - val_loss: 0.6442 - val_acc: 0.7087
Epoch 100/150
514/514 [==============================] - 0s - loss: 0.5013 - acc: 0.7704 - val_loss: 0.6560 - val_acc: 0.7008
Epoch 101/150
514/514 [==============================] - 0s - loss: 0.5459 - acc: 0.7374 - val_loss: 0.6787 - val_acc: 0.6890
Epoch 102/150
514/514 [==============================] - 0s - loss: 0.5184 - acc: 0.7626 - val_loss: 0.6015 - val_acc: 0.7244
Epoch 103/150
514/514 [==============================] - 0s - loss: 0.4902 - acc: 0.7724 - val_loss: 0.6074 - val_acc: 0.7244
Epoch 104/150
514/514 [==============================] - 0s - loss: 0.4989 - acc: 0.7763 - val_loss: 0.6312 - val_acc: 0.7126
Epoch 105/150
514/514 [==============================] - 0s - loss: 0.5064 - acc: 0.7549 - val_loss: 0.6232 - val_acc: 0.7205
Epoch 106/150
514/514 [==============================] - 0s - loss: 0.4992 - acc: 0.7588 - val_loss: 0.6248 - val_acc: 0.7283
Epoch 107/150
514/514 [==============================] - 0s - loss: 0.5326 - acc: 0.7451 - val_loss: 0.6476 - val_acc: 0.6890
Epoch 108/150
514/514 [==============================] - 0s - loss: 0.4882 - acc: 0.7802 - val_loss: 0.6395 - val_acc: 0.6654
Epoch 109/150
514/514 [==============================] - 0s - loss: 0.5056 - acc: 0.7743 - val_loss: 0.5995 - val_acc: 0.7362
Epoch 110/150
514/514 [==============================] - 0s - loss: 0.4828 - acc: 0.7782 - val_loss: 0.6349 - val_acc: 0.7087
Epoch 111/150
514/514 [==============================] - 0s - loss: 0.4801 - acc: 0.7821 - val_loss: 0.6462 - val_acc: 0.6732
Epoch 112/150
514/514 [==============================] - 0s - loss: 0.4938 - acc: 0.7549 - val_loss: 0.6039 - val_acc: 0.7244
Epoch 113/150
514/514 [==============================] - 0s - loss: 0.5066 - acc: 0.7802 - val_loss: 0.6126 - val_acc: 0.7283
Epoch 114/150
514/514 [==============================] - 0s - loss: 0.5020 - acc: 0.7490 - val_loss: 0.6032 - val_acc: 0.7283
Epoch 115/150
514/514 [==============================] - 0s - loss: 0.4987 - acc: 0.7549 - val_loss: 0.6063 - val_acc: 0.7165
Epoch 116/150
514/514 [==============================] - 0s - loss: 0.5207 - acc: 0.7646 - val_loss: 0.6163 - val_acc: 0.7087
Epoch 117/150
514/514 [==============================] - 0s - loss: 0.4993 - acc: 0.7471 - val_loss: 0.6008 - val_acc: 0.7283
Epoch 118/150
514/514 [==============================] - 0s - loss: 0.4954 - acc: 0.7685 - val_loss: 0.5984 - val_acc: 0.7362
Epoch 119/150
514/514 [==============================] - 0s - loss: 0.4874 - acc: 0.7743 - val_loss: 0.6070 - val_acc: 0.7362
Epoch 120/150
514/514 [==============================] - 0s - loss: 0.4909 - acc: 0.7568 - val_loss: 0.5990 - val_acc: 0.7402
Epoch 121/150
514/514 [==============================] - 0s - loss: 0.4781 - acc: 0.7588 - val_loss: 0.6695 - val_acc: 0.6890
Epoch 122/150
514/514 [==============================] - 0s - loss: 0.4861 - acc: 0.7665 - val_loss: 0.6198 - val_acc: 0.7402
Epoch 123/150
514/514 [==============================] - 0s - loss: 0.4786 - acc: 0.7626 - val_loss: 0.6440 - val_acc: 0.6811
Epoch 124/150
514/514 [==============================] - 0s - loss: 0.4830 - acc: 0.7665 - val_loss: 0.6391 - val_acc: 0.6811
Epoch 125/150
514/514 [==============================] - 0s - loss: 0.4707 - acc: 0.7685 - val_loss: 0.6373 - val_acc: 0.7205
Epoch 126/150
514/514 [==============================] - 0s - loss: 0.4953 - acc: 0.7802 - val_loss: 0.5981 - val_acc: 0.7441
Epoch 127/150
514/514 [==============================] - 0s - loss: 0.4689 - acc: 0.7957 - val_loss: 0.5971 - val_acc: 0.7362
Epoch 128/150
514/514 [==============================] - 0s - loss: 0.4938 - acc: 0.7685 - val_loss: 0.8264 - val_acc: 0.6850
Epoch 129/150
514/514 [==============================] - 0s - loss: 0.5091 - acc: 0.7626 - val_loss: 0.6702 - val_acc: 0.6772
Epoch 130/150
514/514 [==============================] - 0s - loss: 0.4939 - acc: 0.7665 - val_loss: 0.6371 - val_acc: 0.7205
Epoch 131/150
514/514 [==============================] - 0s - loss: 0.4974 - acc: 0.7607 - val_loss: 0.6433 - val_acc: 0.6890
Epoch 132/150
514/514 [==============================] - 0s - loss: 0.4957 - acc: 0.7743 - val_loss: 0.7286 - val_acc: 0.6772
Epoch 133/150
514/514 [==============================] - 0s - loss: 0.5178 - acc: 0.7743 - val_loss: 0.6052 - val_acc: 0.7323
Epoch 134/150
514/514 [==============================] - 0s - loss: 0.4986 - acc: 0.7626 - val_loss: 0.5918 - val_acc: 0.7520
Epoch 135/150
514/514 [==============================] - 0s - loss: 0.4705 - acc: 0.7704 - val_loss: 0.5899 - val_acc: 0.7441
Epoch 136/150
514/514 [==============================] - 0s - loss: 0.5097 - acc: 0.7724 - val_loss: 0.6217 - val_acc: 0.7126
Epoch 137/150
514/514 [==============================] - 0s - loss: 0.4939 - acc: 0.7704 - val_loss: 0.6085 - val_acc: 0.7402
Epoch 138/150
514/514 [==============================] - 0s - loss: 0.4722 - acc: 0.7626 - val_loss: 0.6014 - val_acc: 0.7402
Epoch 139/150
514/514 [==============================] - 0s - loss: 0.4921 - acc: 0.7626 - val_loss: 0.5945 - val_acc: 0.7362
Epoch 140/150
514/514 [==============================] - 0s - loss: 0.4802 - acc: 0.7802 - val_loss: 0.6175 - val_acc: 0.7047
Epoch 141/150
514/514 [==============================] - 0s - loss: 0.4786 - acc: 0.7685 - val_loss: 0.6166 - val_acc: 0.7244
Epoch 142/150
514/514 [==============================] - 0s - loss: 0.4773 - acc: 0.7529 - val_loss: 0.5986 - val_acc: 0.7283
Epoch 143/150
514/514 [==============================] - 0s - loss: 0.4803 - acc: 0.7626 - val_loss: 0.6455 - val_acc: 0.7205
Epoch 144/150
514/514 [==============================] - 0s - loss: 0.4776 - acc: 0.7607 - val_loss: 0.5955 - val_acc: 0.7362
Epoch 145/150
514/514 [==============================] - 0s - loss: 0.4808 - acc: 0.7626 - val_loss: 0.6175 - val_acc: 0.7323
Epoch 146/150
514/514 [==============================] - 0s - loss: 0.4659 - acc: 0.7743 - val_loss: 0.6024 - val_acc: 0.7362
Epoch 147/150
514/514 [==============================] - 0s - loss: 0.4818 - acc: 0.7840 - val_loss: 0.6829 - val_acc: 0.7008
Epoch 148/150
514/514 [==============================] - 0s - loss: 0.5261 - acc: 0.7451 - val_loss: 0.6686 - val_acc: 0.7126
Epoch 149/150
514/514 [==============================] - 0s - loss: 0.4856 - acc: 0.7763 - val_loss: 0.6530 - val_acc: 0.7244
Epoch 150/150
514/514 [==============================] - 0s - loss: 0.4944 - acc: 0.7918 - val_loss: 0.6109 - val_acc: 0.7402










    
Out[12]:








<keras.callbacks.History at 0x12059ac18>

Cross Validation





In [13]:



    


from keras.models import Sequential
from keras.layers import Dense
from sklearn.model_selection import StratifiedKFold
import numpy as np



    











In [14]:



    


seed = 7
np.random.seed(seed)



    











In [22]:



    


dataset = np.loadtxt('pima-indians-diabetes.data', delimiter=',')
X = dataset[:, 0:8]
Y = dataset[:, 8]
X.shape



    


















    
Out[22]:








(768, 8)

















In [16]:



    


kfold = StratifiedKFold(n_splits=5, shuffle=True, random_state=seed)



    











In [24]:



    


cvscores = []
for train, test in kfold.split(X, Y):
    model = Sequential()
    model.add(Dense(12, input_dim=8, activation='relu'))
    model.add(Dense(8, activation='relu'))
    model.add(Dense(1, activation='sigmoid'))
    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
    model.fit(X[train], Y[train], epochs=150, batch_size=10, verbose=0)
    scores = model.evaluate(X[test], Y[test], verbose=0)
    print('%s: %.2f%%' % (model.metrics_names[1], scores[1] * 100))
    cvscores.append(scores[1] * 100)
print('%.2f%% (+/- %.2f%%' % (np.mean(cvscores), np.std(cvscores)))



    


















    






acc: 72.73%
acc: 74.03%
acc: 75.97%
acc: 65.36%
acc: 68.63%
71.34% (+/- 3.84%

















In [25]:



    


X.shape



    


















    
Out[25]:








(768, 8)

















In [26]:



    


Y.shape



    


















    
Out[26]:








(768,)

















In [ ]:

Content source: aidiary/notebooks

Similar notebooks:

notebook.community | gallery | about