MNIST练习
时间:2022-07-23
本文章向大家介绍MNIST练习，主要内容包括其使用实例、应用技巧、基本知识点总结和需要注意事项，具有一定的参考价值，需要的朋友可以参考一下。
MNIST 识别手写数字练习
导入包、数据
from keras.datasets import mnist
import matplotlib.pyplot as plt
(X_train, Y_train),(X_test, Y_test) = mnist.load_data()
from keras import Model, Input, metrics, optimizers
from keras.layers import Dense, Conv2D, Activation, MaxPool2D
import keras
import numpy as np训练数据可视化
def mnist_show(X_train, Y_train):
n = 3
m = 3
plt.figure()
for i in range(m):
for j in range(n):
idx = np.random.choice(X_train.shape[0])
plt.subplot(m,n,i*m+j+1)
plt.imshow(X_train[idx])
plt.title(Y_train[idx])
plt.show()
mnist_show(X_train, Y_train)数据预处理
# 数据预处理
X_train = X_train[0:500]
Y_train = Y_train[0:500]
X_test = X_test[0:100]
Y_test = Y_test[0:100]
print(X_train.shape,Y_train.shape)
X_tr = X_train / 255.0
X_te = X_test / 255.0
X_tr = X_tr.reshape(X_tr.shape[0],28,28,1)
X_te = X_te.reshape(X_te.shape[0],28,28,1)
Y_tr = Y_train.reshape(Y_train.shape[0],1)
Y_te = Y_test.reshape(Y_test.shape[0],1)
print(X_tr.shape)
print(X_te.shape)
Y_tr = keras.utils.to_categorical(Y_train, 10)
Y_te = keras.utils.to_categorical(Y_test, 10)
print(Y_tr.shape,Y_te.shape)(500, 28, 28) (500,)
(500, 28, 28, 1)
(100, 28, 28, 1)
(500, 10) (100, 10)定义模型
def model():
inputs = Input(shape=(28,28,1))
x = Conv2D(filters=16,kernel_size=(2,2),strides=1,padding='same')(inputs)
x = keras.layers.BatchNormalization(axis = 3)(x)
x = keras.layers.MaxPool2D(pool_size=(2,2))(x)
x = Conv2D(filters=32,kernel_size=(2,2),strides=1,padding='same')(inputs)
x = keras.layers.BatchNormalization(axis = 3)(x)
x = keras.layers.MaxPool2D(pool_size=(2,2))(x)
x = keras.layers.Flatten()(x)
x = Dense(200,activation='relu')(x)
x = keras.layers.Dropout(0.3)(x)
x = Dense(50,activation='relu')(x)
x = keras.layers.Dropout(0.3)(x)
x = Dense(10,activation='softmax')(x)
model = Model(inputs = inputs, outputs = x)
optimizer = optimizers.Adam(1e-3)
loss = keras.losses.categorical_crossentropy
model.compile(optimizer=optimizer, loss = loss)
model.summary()
return model
mnist_classification = model()模型框图可视化
Model: "model_13"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
input_18 (InputLayer) (None, 28, 28, 1) 0
_________________________________________________________________
conv2d_21 (Conv2D) (None, 28, 28, 32) 160
_________________________________________________________________
batch_normalization_4 (Batch (None, 28, 28, 32) 128
_________________________________________________________________
max_pooling2d_21 (MaxPooling (None, 14, 14, 32) 0
_________________________________________________________________
flatten_16 (Flatten) (None, 6272) 0
_________________________________________________________________
dense_32 (Dense) (None, 200) 1254600
_________________________________________________________________
dropout_10 (Dropout) (None, 200) 0
_________________________________________________________________
dense_33 (Dense) (None, 50) 10050
_________________________________________________________________
dropout_11 (Dropout) (None, 50) 0
_________________________________________________________________
dense_34 (Dense) (None, 10) 510
=================================================================
Total params: 1,265,448
Trainable params: 1,265,384
Non-trainable params: 64
_________________________________________________________________训练模型
mnist_classification.fit(X_tr,Y_tr,batch_size=64,epochs=200)Epoch 1/200
500/500 [==============================] - 1s 3ms/step - loss: 2.1675
Epoch 2/200
500/500 [==============================] - 1s 1ms/step - loss: 1.1908
Epoch 3/200
500/500 [==============================] - 1s 1ms/step - loss: 0.7801
Epoch 4/200
500/500 [==============================] - 1s 1ms/step - loss: 0.6275
Epoch 5/200
500/500 [==============================] - 1s 1ms/step - loss: 0.5564
Epoch 6/200
500/500 [==============================] - 1s 1ms/step - loss: 0.4784
Epoch 7/200
500/500 [==============================] - 1s 1ms/step - loss: 0.3887
Epoch 8/200
500/500 [==============================] - 0s 997us/step - loss: 0.3055
Epoch 9/200
500/500 [==============================] - 1s 1ms/step - loss: 0.3098
Epoch 10/200
500/500 [==============================] - 1s 1ms/step - loss: 0.2750
Epoch 11/200
500/500 [==============================] - 1s 1ms/step - loss: 0.2044
Epoch 12/200
500/500 [==============================] - 1s 1ms/step - loss: 0.2084
Epoch 13/200
500/500 [==============================] - 1s 1ms/step - loss: 0.2392
Epoch 14/200
500/500 [==============================] - 1s 1ms/step - loss: 0.1859
Epoch 15/200
500/500 [==============================] - 1s 2ms/step - loss: 0.1709
Epoch 16/200
500/500 [==============================] - 1s 1ms/step - loss: 0.1494
Epoch 17/200
500/500 [==============================] - 1s 1ms/step - loss: 0.1361
Epoch 18/200
500/500 [==============================] - 1s 2ms/step - loss: 0.1330
Epoch 19/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0902
Epoch 20/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0964
Epoch 21/200
500/500 [==============================] - 1s 1ms/step - loss: 0.1014
Epoch 22/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0904
Epoch 23/200
500/500 [==============================] - 1s 1ms/step - loss: 0.1006
Epoch 24/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0909
Epoch 25/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0794
Epoch 26/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0757
Epoch 27/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0488
Epoch 28/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0530
Epoch 29/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0341
Epoch 30/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0488
Epoch 31/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0559
Epoch 32/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0529
Epoch 33/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0795
Epoch 34/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0413
Epoch 35/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0509
Epoch 36/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0478
Epoch 37/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0289
Epoch 38/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0421
Epoch 39/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0433
Epoch 40/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0484
Epoch 41/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0248
Epoch 42/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0426
Epoch 43/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0438
Epoch 44/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0270
Epoch 45/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0249
Epoch 46/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0308
Epoch 47/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0291
Epoch 48/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0462
Epoch 49/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0518
Epoch 50/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0296
Epoch 51/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0421
Epoch 52/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0524
Epoch 53/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0414
Epoch 54/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0392
Epoch 55/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0211
Epoch 56/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0360
Epoch 57/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0218
Epoch 58/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0285
Epoch 59/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0161
Epoch 60/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0190
Epoch 61/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0242
Epoch 62/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0300
Epoch 63/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0190
Epoch 64/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0200
Epoch 65/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0199
Epoch 66/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0225
Epoch 67/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0245
Epoch 68/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0200
Epoch 69/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0182
Epoch 70/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0173
Epoch 71/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0355
Epoch 72/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0267
Epoch 73/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0310
Epoch 74/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0158
Epoch 75/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0365
Epoch 76/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0214
Epoch 77/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0211
Epoch 78/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0126
Epoch 79/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0136
Epoch 80/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0275
Epoch 81/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0276
Epoch 82/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0353
Epoch 83/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0194
Epoch 84/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0123
Epoch 85/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0281
Epoch 86/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0229
Epoch 87/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0092
Epoch 88/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0309
Epoch 89/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0294
Epoch 90/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0284
Epoch 91/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0363
Epoch 92/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0176
Epoch 93/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0234
Epoch 94/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0162
Epoch 95/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0238
Epoch 96/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0110
Epoch 97/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0247
Epoch 98/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0153
Epoch 99/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0309
Epoch 100/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0095
Epoch 101/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0363
Epoch 102/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0250
Epoch 103/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0252
Epoch 104/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0093
Epoch 105/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0280
Epoch 106/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0137
Epoch 107/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0080
Epoch 108/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0096
Epoch 109/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0157
Epoch 110/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0286
Epoch 111/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0152
Epoch 112/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0108
Epoch 113/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0124
Epoch 114/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0155
Epoch 115/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0163
Epoch 116/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0053
Epoch 117/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0041
Epoch 118/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0121
Epoch 119/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0057
Epoch 120/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0053
Epoch 121/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0169
Epoch 122/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0028
Epoch 123/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0136
Epoch 124/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0323
Epoch 125/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0065
Epoch 126/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0084
Epoch 127/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0281
Epoch 128/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0066
Epoch 129/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0087
Epoch 130/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0082
Epoch 131/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0074
Epoch 132/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0184
Epoch 133/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0198
Epoch 134/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0211
Epoch 135/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0158
Epoch 136/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0345
Epoch 137/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0117
Epoch 138/200
500/500 [==============================] - ETA: 0s - loss: 0.008 - 1s 2ms/step - loss: 0.0081
Epoch 139/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0088
Epoch 140/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0141
Epoch 141/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0077
Epoch 142/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0111
Epoch 143/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0343
Epoch 144/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0046
Epoch 145/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0196
Epoch 146/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0103
Epoch 147/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0120
Epoch 148/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0119
Epoch 149/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0088
Epoch 150/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0132
Epoch 151/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0094
Epoch 152/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0147
Epoch 153/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0036
Epoch 154/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0241
Epoch 155/200
500/500 [==============================] - 2s 5ms/step - loss: 0.0234
Epoch 156/200
500/500 [==============================] - 1s 3ms/step - loss: 0.0162
Epoch 157/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0126
Epoch 158/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0228
Epoch 159/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0186
Epoch 160/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0162
Epoch 161/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0144
Epoch 162/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0260
Epoch 163/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0248
Epoch 164/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0116
Epoch 165/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0214
Epoch 166/200
500/500 [==============================] - 3s 6ms/step - loss: 0.0120
Epoch 167/200
500/500 [==============================] - 2s 4ms/step - loss: 0.0263
Epoch 168/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0110
Epoch 169/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0351
Epoch 170/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0164
Epoch 171/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0131
Epoch 172/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0247
Epoch 173/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0234
Epoch 174/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0322
Epoch 175/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0259
Epoch 176/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0089
Epoch 177/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0106
Epoch 178/200
500/500 [==============================] - 1s 2ms/step - loss: 8.5937e-04
Epoch 179/200
500/500 [==============================] - 2s 3ms/step - loss: 0.0069
Epoch 180/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0118
Epoch 181/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0030
Epoch 182/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0084
Epoch 183/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0171
Epoch 184/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0344
Epoch 185/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0123
Epoch 186/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0103
Epoch 187/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0073
Epoch 188/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0337
Epoch 189/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0174
Epoch 190/200
500/500 [==============================] - 1s 2ms/step - loss: 0.0112
Epoch 191/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0131
Epoch 192/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0016
Epoch 193/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0112
Epoch 194/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0179
Epoch 195/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0119
Epoch 196/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0150
Epoch 197/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0101
Epoch 198/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0018
Epoch 199/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0030
Epoch 200/200
500/500 [==============================] - 1s 1ms/step - loss: 0.0161
<keras.callbacks.callbacks.History at 0x1356fe890>评价模型
score = mnist_classification.evaluate(X_te,Y_te,verbose=1)
print(score)100/100 [==============================] - 0s 2ms/step
0.7463939571380616Y_pre = mnist_classification.predict(X_te)
print(type(Y_pre))<class 'numpy.ndarray'>Y_p_val = np.argmax(Y_pre,axis=-1)print(Y_test[0:100]-Y_p_val)[ 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 2 0 0 1 0 0 0 0 0
0 0 -2 0 0 0 0 0 0 -2 0 0 0 0 0 0 0 0 0 0 0 0 -4 0
0 0 0 0 0 0 0 0 0 0 0 0 0 6 0 0 4 -5 4 0 0 0 0 0
0 1 0 0 0 0 0 0 -2 0 0 -2 0 0 0 0 0 0 0 0 0 0 0 0
-7 0 0 0]import keras_lr_finderlr_finder = keras_lr_finder.LRFinder(mnist_classification)
lr_finder.find(X_tr,Y_tr,start_lr = 1e-5, end_lr = 1e2,batch_size = 200, epochs = 10)Epoch 1/10
500/500 [==============================] - 1s 2ms/step - loss: 0.0054
Epoch 2/10
400/500 [=======================>......] - ETA: 0s - loss: 0.0047lr_finder.plot_loss(n_skip_beginning=1,n_skip_end=1)存取模型
n
mnist_classification.save(filepath='mnist_cla.h5',include_optimizer='Adam',overwrite=True)aa = keras.models.load_model('mnist_cla.h5')aa.summary()Model: "model_13"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
input_18 (InputLayer) (None, 28, 28, 1) 0
_________________________________________________________________
conv2d_21 (Conv2D) (None, 28, 28, 32) 160
_________________________________________________________________
batch_normalization_4 (Batch (None, 28, 28, 32) 128
_________________________________________________________________
max_pooling2d_21 (MaxPooling (None, 14, 14, 32) 0
_________________________________________________________________
flatten_16 (Flatten) (None, 6272) 0
_________________________________________________________________
dense_32 (Dense) (None, 200) 1254600
_________________________________________________________________
dropout_10 (Dropout) (None, 200) 0
_________________________________________________________________
dense_33 (Dense) (None, 50) 10050
_________________________________________________________________
dropout_11 (Dropout) (None, 50) 0
_________________________________________________________________
dense_34 (Dense) (None, 10) 510
=================================================================
Total params: 1,265,448
Trainable params: 1,265,384
Non-trainable params: 64
_________________________________________________________________
- 通过top命令抓取cpu高消耗的sql （44天）
- 关于字符串匹配查找的总结（43天)
- 一条sql语句导致的数据库宕机问题及分析（42天)
- 外部表的导入导出问题 （41天)
- 当我们和计算机交互时，它看到的是什么？
- 一条sql语句“导致”的数据库宕机问题及分析 （38天)
- rman数据备份恢复学习笔记(49天)
- 虚拟专用数据库VPD应用 (48天)
- 关于创建视图的问题(48天)
- 性能调优之redo切换频率(47天)
- 关于oracle中session跟踪的总结(56天)
- oracle中关于小数中0的格式化（55天)
- 关于trigger过滤最大值的问题(54天)
- oracle共享服务器配置汇总(53天)
- JavaScript 教程
- JavaScript 编辑工具
- JavaScript 与HTML
- JavaScript 与Java
- JavaScript 数据结构
- JavaScript 基本数据类型
- JavaScript 特殊数据类型
- JavaScript 运算符
- JavaScript typeof 运算符
- JavaScript 表达式
- JavaScript 类型转换
- JavaScript 基本语法
- JavaScript 注释
- Javascript 基本处理流程
- Javascript 选择结构
- Javascript if 语句
- Javascript if 语句的嵌套
- Javascript switch 语句
- Javascript 循环结构
- Javascript 循环结构实例
- Javascript 跳转语句
- Javascript 控制语句总结
- Javascript 函数介绍
- Javascript 函数的定义
- Javascript 函数调用
- Javascript 几种特殊的函数
- JavaScript 内置函数简介
- Javascript eval() 函数
- Javascript isFinite() 函数
- Javascript isNaN() 函数
- parseInt() 与 parseFloat()
- escape() 与 unescape()
- Javascript 字符串介绍
- Javascript length属性
- javascript 字符串函数
- Javascript 日期对象简介
- Javascript 日期对象用途
- Date 对象属性和方法
- Javascript 数组是什么
- Javascript 创建数组
- Javascript 数组赋值与取值
- Javascript 数组属性和方法
- keras自定义损失函数并且模型加载的写法介绍
- pandas DataFrame运算的实现
- Python流程控制语句的深入讲解
- 在keras里面实现计算f1-score的代码
- Keras官方中文文档:性能评估Metrices详解
- Django QuerySet查询集原理及代码实例
- Python中zipfile压缩文件模块的基本使用教程
- 基于nexus3配置Python仓库过程详解
- Python Django中间件使用原理及流程分析
- keras读取h5文件load_weights、load代码操作
- 简单了解Python变量作用域正确使用方法
- Python实现Keras搭建神经网络训练分类模型教程
- keras导入weights方式
- 使用Keras实现简单线性回归模型操作
- 音频处理 windows10下python三方库librosa安装教程