其他分享
首页 > 其他分享> > keras网络模型

keras网络模型

作者:互联网

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

# Generate dummy data
x_train = np.random.random((1000, 20))
y_train = np.random.randint(2, size=(1000, 1))
x_test = np.random.random((100, 20))
y_test = np.random.randint(2, size=(100, 1))

model = Sequential()
model.add(Dense(64, input_dim=20, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(64, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(1, activation='sigmoid'))

model.compile(loss='binary_crossentropy',
              optimizer='rmsprop',
              metrics=['accuracy'])
model.fit(x_train, y_train,
          epochs=200,
          batch_size=128)
score = model.evaluate(x_test, y_test, batch_size=128)
print(keras.__version__)
model.summary()
Epoch 1/200
1000/1000 [==============================] - 0s - loss: 0.7169 - acc: 0.4930     
Epoch 2/200
1000/1000 [==============================] - 0s - loss: 0.7091 - acc: 0.5000     
Epoch 3/200
1000/1000 [==============================] - 0s - loss: 0.7099 - acc: 0.5120     
Epoch 4/200
1000/1000 [==============================] - 0s - loss: 0.7016 - acc: 0.4940     
Epoch 5/200
1000/1000 [==============================] - 0s - loss: 0.7037 - acc: 0.5010     
Epoch 6/200
1000/1000 [==============================] - 0s - loss: 0.6971 - acc: 0.5060     
Epoch 7/200
1000/1000 [==============================] - 0s - loss: 0.7024 - acc: 0.4880     
Epoch 8/200
1000/1000 [==============================] - 0s - loss: 0.6945 - acc: 0.5130     
Epoch 9/200
1000/1000 [==============================] - 0s - loss: 0.6963 - acc: 0.4990     
Epoch 10/200
1000/1000 [==============================] - 0s - loss: 0.6980 - acc: 0.5000     
Epoch 11/200
1000/1000 [==============================] - 0s - loss: 0.6869 - acc: 0.5500     
Epoch 12/200
1000/1000 [==============================] - 0s - loss: 0.6936 - acc: 0.5140     
Epoch 13/200
1000/1000 [==============================] - 0s - loss: 0.6959 - acc: 0.5080     
Epoch 14/200
1000/1000 [==============================] - 0s - loss: 0.7027 - acc: 0.4810     
Epoch 15/200
1000/1000 [==============================] - 0s - loss: 0.6911 - acc: 0.5200     
Epoch 16/200
1000/1000 [==============================] - 0s - loss: 0.6879 - acc: 0.5210     
Epoch 17/200
1000/1000 [==============================] - 0s - loss: 0.6988 - acc: 0.5000     
Epoch 18/200
1000/1000 [==============================] - 0s - loss: 0.6972 - acc: 0.4990     
Epoch 19/200
1000/1000 [==============================] - 0s - loss: 0.6929 - acc: 0.5380     
Epoch 20/200
1000/1000 [==============================] - 0s - loss: 0.6955 - acc: 0.5090     
Epoch 21/200
1000/1000 [==============================] - 0s - loss: 0.6897 - acc: 0.5200     
Epoch 22/200
1000/1000 [==============================] - 0s - loss: 0.6856 - acc: 0.5360     
Epoch 23/200
1000/1000 [==============================] - 0s - loss: 0.6860 - acc: 0.5510     
Epoch 24/200
1000/1000 [==============================] - 0s - loss: 0.6947 - acc: 0.5030     
Epoch 25/200
1000/1000 [==============================] - 0s - loss: 0.6913 - acc: 0.5280     
Epoch 26/200
1000/1000 [==============================] - 0s - loss: 0.6828 - acc: 0.5470     
Epoch 27/200
1000/1000 [==============================] - 0s - loss: 0.6876 - acc: 0.5370     
Epoch 28/200
1000/1000 [==============================] - 0s - loss: 0.6872 - acc: 0.5360     
Epoch 29/200
1000/1000 [==============================] - 0s - loss: 0.6926 - acc: 0.5250     
Epoch 30/200
1000/1000 [==============================] - 0s - loss: 0.6878 - acc: 0.5380     
Epoch 31/200
1000/1000 [==============================] - 0s - loss: 0.6889 - acc: 0.5420     
Epoch 32/200
1000/1000 [==============================] - 0s - loss: 0.6885 - acc: 0.5350     
Epoch 33/200
1000/1000 [==============================] - 0s - loss: 0.6858 - acc: 0.5470     
Epoch 34/200
1000/1000 [==============================] - 0s - loss: 0.6902 - acc: 0.5410     
Epoch 35/200
1000/1000 [==============================] - 0s - loss: 0.6837 - acc: 0.5620     
Epoch 36/200
1000/1000 [==============================] - 0s - loss: 0.6854 - acc: 0.5430     
Epoch 37/200
1000/1000 [==============================] - 0s - loss: 0.6867 - acc: 0.5220     
Epoch 38/200
1000/1000 [==============================] - 0s - loss: 0.6830 - acc: 0.5490     
Epoch 39/200
1000/1000 [==============================] - 0s - loss: 0.6829 - acc: 0.5510     
Epoch 40/200
1000/1000 [==============================] - 0s - loss: 0.6826 - acc: 0.5630     
Epoch 41/200
1000/1000 [==============================] - 0s - loss: 0.6869 - acc: 0.5320     
Epoch 42/200
1000/1000 [==============================] - 0s - loss: 0.6849 - acc: 0.5440     
Epoch 43/200
1000/1000 [==============================] - 0s - loss: 0.6855 - acc: 0.5430     
Epoch 44/200
1000/1000 [==============================] - 0s - loss: 0.6797 - acc: 0.5550     
Epoch 45/200
1000/1000 [==============================] - 0s - loss: 0.6795 - acc: 0.5870     
Epoch 46/200
1000/1000 [==============================] - 0s - loss: 0.6842 - acc: 0.5460     
Epoch 47/200
1000/1000 [==============================] - 0s - loss: 0.6824 - acc: 0.5510     
Epoch 48/200
1000/1000 [==============================] - 0s - loss: 0.6835 - acc: 0.5530     
Epoch 49/200
1000/1000 [==============================] - 0s - loss: 0.6852 - acc: 0.5500     
Epoch 50/200
1000/1000 [==============================] - 0s - loss: 0.6774 - acc: 0.5740     
Epoch 51/200
1000/1000 [==============================] - 0s - loss: 0.6764 - acc: 0.5670     
Epoch 52/200
1000/1000 [==============================] - 0s - loss: 0.6781 - acc: 0.5490     
Epoch 53/200
1000/1000 [==============================] - 0s - loss: 0.6770 - acc: 0.5520     
Epoch 54/200
1000/1000 [==============================] - 0s - loss: 0.6785 - acc: 0.5700     
Epoch 55/200
1000/1000 [==============================] - 0s - loss: 0.6746 - acc: 0.5810     
Epoch 56/200
1000/1000 [==============================] - 0s - loss: 0.6820 - acc: 0.5540     
Epoch 57/200
1000/1000 [==============================] - 0s - loss: 0.6819 - acc: 0.5660     
Epoch 58/200
1000/1000 [==============================] - 0s - loss: 0.6647 - acc: 0.5950     
Epoch 59/200
1000/1000 [==============================] - 0s - loss: 0.6843 - acc: 0.5610     
Epoch 60/200
1000/1000 [==============================] - 0s - loss: 0.6812 - acc: 0.5590     
Epoch 61/200
1000/1000 [==============================] - 0s - loss: 0.6776 - acc: 0.5610     
Epoch 62/200
1000/1000 [==============================] - 0s - loss: 0.6777 - acc: 0.5550     
Epoch 63/200
1000/1000 [==============================] - 0s - loss: 0.6824 - acc: 0.5510     
Epoch 64/200
1000/1000 [==============================] - 0s - loss: 0.6740 - acc: 0.5860     
Epoch 65/200
1000/1000 [==============================] - 0s - loss: 0.6710 - acc: 0.5870     
Epoch 66/200
1000/1000 [==============================] - 0s - loss: 0.6706 - acc: 0.5860     
Epoch 67/200
1000/1000 [==============================] - 0s - loss: 0.6714 - acc: 0.5730     
Epoch 68/200
1000/1000 [==============================] - 0s - loss: 0.6695 - acc: 0.5870     
Epoch 69/200
1000/1000 [==============================] - 0s - loss: 0.6703 - acc: 0.5890     
Epoch 70/200
1000/1000 [==============================] - 0s - loss: 0.6732 - acc: 0.5620     
Epoch 71/200
1000/1000 [==============================] - 0s - loss: 0.6749 - acc: 0.5630     
Epoch 72/200
1000/1000 [==============================] - 0s - loss: 0.6661 - acc: 0.5890     
Epoch 73/200
1000/1000 [==============================] - 0s - loss: 0.6687 - acc: 0.5990     
Epoch 74/200
1000/1000 [==============================] - 0s - loss: 0.6721 - acc: 0.5910     
Epoch 75/200
1000/1000 [==============================] - 0s - loss: 0.6632 - acc: 0.5830     
Epoch 76/200
1000/1000 [==============================] - 0s - loss: 0.6662 - acc: 0.6020     
Epoch 77/200
1000/1000 [==============================] - 0s - loss: 0.6718 - acc: 0.5690     
Epoch 78/200
1000/1000 [==============================] - 0s - loss: 0.6717 - acc: 0.5670     
Epoch 79/200
1000/1000 [==============================] - 0s - loss: 0.6658 - acc: 0.5890     
Epoch 80/200
1000/1000 [==============================] - 0s - loss: 0.6672 - acc: 0.5750     
Epoch 81/200
1000/1000 [==============================] - 0s - loss: 0.6589 - acc: 0.6020     
Epoch 82/200
1000/1000 [==============================] - 0s - loss: 0.6699 - acc: 0.5740     
Epoch 83/200
1000/1000 [==============================] - 0s - loss: 0.6626 - acc: 0.5900     
Epoch 84/200
1000/1000 [==============================] - 0s - loss: 0.6648 - acc: 0.5960     
Epoch 85/200
1000/1000 [==============================] - 0s - loss: 0.6716 - acc: 0.5760     
Epoch 86/200
1000/1000 [==============================] - 0s - loss: 0.6594 - acc: 0.5950     
Epoch 87/200
1000/1000 [==============================] - 0s - loss: 0.6603 - acc: 0.5830     
Epoch 88/200
1000/1000 [==============================] - 0s - loss: 0.6627 - acc: 0.5940     
Epoch 89/200
1000/1000 [==============================] - 0s - loss: 0.6631 - acc: 0.6060     
Epoch 90/200
1000/1000 [==============================] - 0s - loss: 0.6606 - acc: 0.5980     
Epoch 91/200
1000/1000 [==============================] - 0s - loss: 0.6534 - acc: 0.6060     
Epoch 92/200
1000/1000 [==============================] - 0s - loss: 0.6706 - acc: 0.5720     
Epoch 93/200
1000/1000 [==============================] - 0s - loss: 0.6455 - acc: 0.6340     
Epoch 94/200
1000/1000 [==============================] - 0s - loss: 0.6592 - acc: 0.5990     
Epoch 95/200
1000/1000 [==============================] - 0s - loss: 0.6554 - acc: 0.5920     
Epoch 96/200
1000/1000 [==============================] - 0s - loss: 0.6606 - acc: 0.5930     
Epoch 97/200
1000/1000 [==============================] - 0s - loss: 0.6570 - acc: 0.6070     
Epoch 98/200
1000/1000 [==============================] - 0s - loss: 0.6503 - acc: 0.6250     
Epoch 99/200
1000/1000 [==============================] - 0s - loss: 0.6505 - acc: 0.6080     
Epoch 100/200
1000/1000 [==============================] - 0s - loss: 0.6534 - acc: 0.6160     
Epoch 101/200
1000/1000 [==============================] - 0s - loss: 0.6535 - acc: 0.6190     
Epoch 102/200
1000/1000 [==============================] - 0s - loss: 0.6595 - acc: 0.6000     
Epoch 103/200
1000/1000 [==============================] - 0s - loss: 0.6616 - acc: 0.5970     
Epoch 104/200
1000/1000 [==============================] - 0s - loss: 0.6469 - acc: 0.6120     
Epoch 105/200
1000/1000 [==============================] - 0s - loss: 0.6526 - acc: 0.6060     
Epoch 106/200
1000/1000 [==============================] - 0s - loss: 0.6544 - acc: 0.5920     
Epoch 107/200
1000/1000 [==============================] - 0s - loss: 0.6523 - acc: 0.6140     
Epoch 108/200
1000/1000 [==============================] - 0s - loss: 0.6465 - acc: 0.6230     
Epoch 109/200
1000/1000 [==============================] - 0s - loss: 0.6417 - acc: 0.6250     
Epoch 110/200
1000/1000 [==============================] - 0s - loss: 0.6447 - acc: 0.6160     
Epoch 111/200
1000/1000 [==============================] - 0s - loss: 0.6484 - acc: 0.6130     
Epoch 112/200
1000/1000 [==============================] - 0s - loss: 0.6470 - acc: 0.6030     
Epoch 113/200
1000/1000 [==============================] - 0s - loss: 0.6537 - acc: 0.5960     
Epoch 114/200
1000/1000 [==============================] - 0s - loss: 0.6458 - acc: 0.6190     
Epoch 115/200
1000/1000 [==============================] - 0s - loss: 0.6395 - acc: 0.6110     
Epoch 116/200
1000/1000 [==============================] - 0s - loss: 0.6499 - acc: 0.6070     
Epoch 117/200
1000/1000 [==============================] - 0s - loss: 0.6320 - acc: 0.6340     
Epoch 118/200
1000/1000 [==============================] - 0s - loss: 0.6390 - acc: 0.6120     
Epoch 119/200
1000/1000 [==============================] - 0s - loss: 0.6394 - acc: 0.6240     
Epoch 120/200
1000/1000 [==============================] - 0s - loss: 0.6503 - acc: 0.6280     
Epoch 121/200
1000/1000 [==============================] - 0s - loss: 0.6409 - acc: 0.6390     
Epoch 122/200
1000/1000 [==============================] - 0s - loss: 0.6463 - acc: 0.6060     
Epoch 123/200
1000/1000 [==============================] - 0s - loss: 0.6351 - acc: 0.6170     
Epoch 124/200
1000/1000 [==============================] - 0s - loss: 0.6334 - acc: 0.6410     
Epoch 125/200
1000/1000 [==============================] - 0s - loss: 0.6446 - acc: 0.6190     
Epoch 126/200
1000/1000 [==============================] - 0s - loss: 0.6335 - acc: 0.6200     
Epoch 127/200
1000/1000 [==============================] - 0s - loss: 0.6342 - acc: 0.6420     
Epoch 128/200
1000/1000 [==============================] - 0s - loss: 0.6349 - acc: 0.6260     
Epoch 129/200
1000/1000 [==============================] - 0s - loss: 0.6362 - acc: 0.6090     
Epoch 130/200
1000/1000 [==============================] - 0s - loss: 0.6385 - acc: 0.6340     
Epoch 131/200
1000/1000 [==============================] - 0s - loss: 0.6286 - acc: 0.6490     
Epoch 132/200
1000/1000 [==============================] - 0s - loss: 0.6292 - acc: 0.6400     
Epoch 133/200
1000/1000 [==============================] - 0s - loss: 0.6394 - acc: 0.6350     
Epoch 134/200
1000/1000 [==============================] - 0s - loss: 0.6335 - acc: 0.6480     
Epoch 135/200
1000/1000 [==============================] - 0s - loss: 0.6302 - acc: 0.6380     
Epoch 136/200
1000/1000 [==============================] - 0s - loss: 0.6206 - acc: 0.6520     
Epoch 137/200
1000/1000 [==============================] - 0s - loss: 0.6252 - acc: 0.6600     
Epoch 138/200
1000/1000 [==============================] - 0s - loss: 0.6241 - acc: 0.6410     
Epoch 139/200
1000/1000 [==============================] - 0s - loss: 0.6315 - acc: 0.6250     
Epoch 140/200
1000/1000 [==============================] - 0s - loss: 0.6207 - acc: 0.6450     
Epoch 141/200
1000/1000 [==============================] - 0s - loss: 0.6258 - acc: 0.6420     
Epoch 142/200
1000/1000 [==============================] - 0s - loss: 0.6350 - acc: 0.6240     
Epoch 143/200
1000/1000 [==============================] - 0s - loss: 0.6265 - acc: 0.6380     
Epoch 144/200
1000/1000 [==============================] - 0s - loss: 0.6278 - acc: 0.6400     
Epoch 145/200
1000/1000 [==============================] - 0s - loss: 0.6339 - acc: 0.6320     
Epoch 146/200
1000/1000 [==============================] - 0s - loss: 0.6207 - acc: 0.6640     
Epoch 147/200
1000/1000 [==============================] - 0s - loss: 0.6319 - acc: 0.6400     
Epoch 148/200
1000/1000 [==============================] - 0s - loss: 0.6102 - acc: 0.6320     
Epoch 149/200
1000/1000 [==============================] - 0s - loss: 0.6303 - acc: 0.6260     
Epoch 150/200
1000/1000 [==============================] - 0s - loss: 0.6209 - acc: 0.6360     
Epoch 151/200
1000/1000 [==============================] - 0s - loss: 0.6112 - acc: 0.6480     
Epoch 152/200
1000/1000 [==============================] - 0s - loss: 0.6315 - acc: 0.6320     
Epoch 153/200
1000/1000 [==============================] - 0s - loss: 0.6174 - acc: 0.6480     
Epoch 154/200
1000/1000 [==============================] - 0s - loss: 0.6222 - acc: 0.6340     
Epoch 155/200
1000/1000 [==============================] - 0s - loss: 0.6284 - acc: 0.6270     
Epoch 156/200
1000/1000 [==============================] - 0s - loss: 0.6251 - acc: 0.6470     
Epoch 157/200
1000/1000 [==============================] - 0s - loss: 0.6163 - acc: 0.6570     
Epoch 158/200
1000/1000 [==============================] - 0s - loss: 0.6119 - acc: 0.6600     
Epoch 159/200
1000/1000 [==============================] - 0s - loss: 0.6227 - acc: 0.6400     
Epoch 160/200
1000/1000 [==============================] - 0s - loss: 0.6178 - acc: 0.6540     
Epoch 161/200
1000/1000 [==============================] - 0s - loss: 0.5987 - acc: 0.6700     
Epoch 162/200
1000/1000 [==============================] - 0s - loss: 0.6195 - acc: 0.6520     
Epoch 163/200
1000/1000 [==============================] - 0s - loss: 0.5964 - acc: 0.6770     
Epoch 164/200
1000/1000 [==============================] - 0s - loss: 0.6135 - acc: 0.6360     
Epoch 165/200
1000/1000 [==============================] - 0s - loss: 0.6135 - acc: 0.6600     
Epoch 166/200
1000/1000 [==============================] - 0s - loss: 0.6076 - acc: 0.6680     
Epoch 167/200
1000/1000 [==============================] - 0s - loss: 0.6119 - acc: 0.6470     
Epoch 168/200
1000/1000 [==============================] - 0s - loss: 0.6224 - acc: 0.6330     
Epoch 169/200
1000/1000 [==============================] - 0s - loss: 0.6083 - acc: 0.6570     
Epoch 170/200
1000/1000 [==============================] - 0s - loss: 0.6158 - acc: 0.6310     
Epoch 171/200
1000/1000 [==============================] - 0s - loss: 0.5967 - acc: 0.6670     
Epoch 172/200
1000/1000 [==============================] - 0s - loss: 0.6069 - acc: 0.6620     
Epoch 173/200
1000/1000 [==============================] - 0s - loss: 0.6173 - acc: 0.6520     
Epoch 174/200
1000/1000 [==============================] - 0s - loss: 0.6081 - acc: 0.6560     
Epoch 175/200
1000/1000 [==============================] - 0s - loss: 0.6028 - acc: 0.6750     
Epoch 176/200
1000/1000 [==============================] - 0s - loss: 0.5989 - acc: 0.6760     
Epoch 177/200
1000/1000 [==============================] - 0s - loss: 0.5975 - acc: 0.6690     
Epoch 178/200
1000/1000 [==============================] - 0s - loss: 0.6016 - acc: 0.6820     
Epoch 179/200
1000/1000 [==============================] - 0s - loss: 0.5929 - acc: 0.6660     
Epoch 180/200
1000/1000 [==============================] - 0s - loss: 0.6043 - acc: 0.6740     
Epoch 181/200
1000/1000 [==============================] - 0s - loss: 0.5839 - acc: 0.6730     
Epoch 182/200
1000/1000 [==============================] - 0s - loss: 0.6007 - acc: 0.6650     
Epoch 183/200
1000/1000 [==============================] - 0s - loss: 0.6087 - acc: 0.6680     
Epoch 184/200
1000/1000 [==============================] - 0s - loss: 0.5902 - acc: 0.6910     
Epoch 185/200
1000/1000 [==============================] - 0s - loss: 0.6017 - acc: 0.6660     
Epoch 186/200
1000/1000 [==============================] - 0s - loss: 0.5949 - acc: 0.6640     
Epoch 187/200
1000/1000 [==============================] - 0s - loss: 0.6036 - acc: 0.6590     
Epoch 188/200
1000/1000 [==============================] - 0s - loss: 0.6023 - acc: 0.6570     
Epoch 189/200
1000/1000 [==============================] - 0s - loss: 0.5892 - acc: 0.6710     
Epoch 190/200
1000/1000 [==============================] - 0s - loss: 0.6060 - acc: 0.6570     
Epoch 191/200
1000/1000 [==============================] - 0s - loss: 0.5939 - acc: 0.6690     
Epoch 192/200
1000/1000 [==============================] - 0s - loss: 0.5912 - acc: 0.6970     
Epoch 193/200
1000/1000 [==============================] - 0s - loss: 0.5863 - acc: 0.6870     
Epoch 194/200
1000/1000 [==============================] - 0s - loss: 0.5917 - acc: 0.6630     
Epoch 195/200
1000/1000 [==============================] - 0s - loss: 0.5739 - acc: 0.6960     
Epoch 196/200
1000/1000 [==============================] - 0s - loss: 0.5986 - acc: 0.6780     
Epoch 197/200
1000/1000 [==============================] - 0s - loss: 0.6069 - acc: 0.6570     
Epoch 198/200
1000/1000 [==============================] - 0s - loss: 0.5793 - acc: 0.6860     
Epoch 199/200
1000/1000 [==============================] - 0s - loss: 0.5962 - acc: 0.6720     
Epoch 200/200
1000/1000 [==============================] - 0s - loss: 0.5961 - acc: 0.6870     
100/100 [==============================] - 0s
2.0.6
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense_33 (Dense)             (None, 64)                1344      
_________________________________________________________________
dropout_17 (Dropout)         (None, 64)                0         
_________________________________________________________________
dense_34 (Dense)             (None, 64)                4160      
_________________________________________________________________
dropout_18 (Dropout)         (None, 64)                0         
_________________________________________________________________
dense_35 (Dense)             (None, 1)                 65        
=================================================================
Total params: 5,569
Trainable params: 5,569
Non-trainable params: 0
_________________________________________________________________

标签:acc,0s,200,模型,keras,网络,Epoch,loss,1000
来源: https://blog.csdn.net/tugouxp/article/details/120403864