3.25日记

一，感知模型的网络架构如下：

运行model.summary()后，你可以看到每一层的详细输出形状和参数数量。以下是部分输出示例：

Model: "sequential"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 conv2d (Conv2D)             (None, 25, 4, 32)         1184      
 max_pooling2d (MaxPooling2D  (None, 13, 2, 32)        0         
_________________________________________________________________
 conv2d_1 (Conv2D)           (None, 13, 2, 64)         18496     
 max_pooling2d_1 (MaxPooling  (None, 7, 1, 64)         0         
_________________________________________________________________
 conv2d_2 (Conv2D)           (None, 7, 1, 128)         73856     
 max_pooling2d_2 (MaxPooling  (None, 4, 1, 128)        0         
_________________________________________________________________
 flatten (Flatten)           (None, 512)               0         
_________________________________________________________________
 dense (Dense)               (None, 128)               65664     
_________________________________________________________________
 dropout (Dropout)           (None, 128)               0         
_________________________________________________________________
 dense_1 (Dense)             (None, 4)                 516       
=================================================================
Total params: 159,712
Trainable params: 159,712
Non-trainable params: 0