Layer (type) Output Shape Param #
=================================================================
conv2d_22 (Conv2D) (None, 6, 52, 2) 52
_________________________________________________________________
activation_64 (Activation) (None, 6, 52, 2) 0
_________________________________________________________________
max_pooling2d_22 (MaxPooling (None, 3, 26, 2) 0
_________________________________________________________________
flatten_22 (Flatten) (None, 156) 0
_________________________________________________________________
dense_43 (Dense) (None, 100) 15700
_________________________________________________________________
activation_65 (Activation) (None, 100) 0
_________________________________________________________________
dense_44 (Dense) (None, 3) 303
_________________________________________________________________
activation_66 (Activation) (None, 3) 0
=================================================================
Total params: 16,055
Trainable params: 16,055
Non-trainable params: 0
conv_layer_bias = [-0.3386867 -0.5459513]
np.max(conv_layer_weight) = 0.6478661894798279
np.min(conv_layer_weight) = -0.44461962580680847
np.max(abs_conv_layer_weight) = 0.6478661894798279
np.min(abs_conv_layer_weight) = 0.01398547738790512
np.max(conv_layer_bias) = -0.3386867046356201
np.min(conv_layer_bias) = -0.5459513068199158
np.max(abs_conv_layer_bias) = 0.5459513068199158
np.min(abs_conv_layer_bias) = 0.3386867046356201
conv_output = (26550, 6, 52, 2)
np.std(conv_output) = 0.2930081784725189
np.max(conv_output) = 1.9703364372253418
np.min(conv_output) = -0.6579744815826416
np.max(abs_conv) = 1.9703364372253418
np.min(abs_conv) = 7.748603820800781e-07
np.max(dence_layer1_weight) = 0.6284171938896179
np.min(dence_layer1_weight) = -0.452088862657547
np.max(abs_dence_layer1_weight) = 0.6284171938896179
np.min(abs_dence_layer1_weight) = 2.8568319976329803e-07
np.max(dence_layer1_bias) = 0.11784756183624268
np.min(dence_layer1_bias) = -0.08143308758735657
np.max(abs_dence_layer1_bias) = 0.11784756183624268
np.min(abs_dence_layer1_bias) = 0.0
dence_layer1_output = (26550, 100)
np.std(dence_layer1_output) = 0.9401099681854248
np.max(dence_layer1_output) = 5.256106853485107
np.min(dence_layer1_output) = -5.39329195022583
np.max(abs_dence_layer1_output) = 5.39329195022583
np.min(abs_dence_layer1_output) = 8.740462362766266e-07
np.max(dence_layer2_weight) = 1.2727510929107666
np.min(dence_layer2_weight) = -1.23631751537323
np.max(abs_dence_layer2_weight) = 1.2727510929107666
np.min(abs_dence_layer2_weight) = 0.002743939170613885
np.max(dence_layer2_bias) = 0.009167053736746311
np.min(dence_layer2_bias) = -0.028219975531101227
np.max(abs_dence_layer2_bias) = 0.028219975531101227
np.min(abs_dence_layer2_bias) = 0.007371054962277412
dence_layer2_output = (26550, 3)
np.std(dence_layer2_output) = 5.037158012390137
np.max(dence_layer2_output) = 17.514923095703125
np.min(dence_layer2_output) = -16.942663192749023
np.max(abs_dence_layer2_output) = 17.514923095703125
np.min(abs_dence_layer2_output) = 0.00031781941652297974
日 | 月 | 火 | 水 | 木 | 金 | 土 |
---|---|---|---|---|---|---|
- | - | - | - | - | 1 | 2 |
3 | 4 | 5 | 6 | 7 | 8 | 9 |
10 | 11 | 12 | 13 | 14 | 15 | 16 |
17 | 18 | 19 | 20 | 21 | 22 | 23 |
24 | 25 | 26 | 27 | 28 | 29 | 30 |
31 | - | - | - | - | - | - |