Table 2 Summary of the optimized neural network architecture with layers, output shape (OS), number of parameters (NP), activation function (AF), number of filters (NF), kernel size (KS), and number of neurons (NN).

First hidden convolutional layer (1D)

(None, 103, 90)

360

relu

90

3

× 

Second hidden convolutional layer (1D)

(None, 103, 70)

18,970

relu

70

3

× 

Maximum pooling layer

(None, 51, 70)

0

× 

× 

× 

× 

Flatten layer

(None, 3570)

0

× 

× 

× 

× 

First hidden dense layer

(None, 512)

1,828,352

relu

× 

× 

512

Second hidden dense layer

(None, 512)

262,656

relu

× 

× 

512

Third hidden dense layer

(None, 512)

262,656

relu

× 

× 

512

Dense output layer

(None, 13)

6669

Linear

× 

× 

13