Atrial Fibrillation Beat Identification Using the Combination of Modified Frequency Slice Wavelet Transform and Convolutional Neural Networks

Table 3

The optimal CNN specifications designed for the ECG classification problem.

Parameters	Values
Learning rate	0.001
First convolutional layer kernel size	10  9
Number of feature maps in the first convolutional and subsampling layer	32
Second convolutional layer kernel size	8  7
Number of feature maps in the second convolutional and subsampling layer	20
Third convolutional layer kernel size	9
Number of feature maps in the third convolutional and subsampling layer	16
Subsampling layer kernel size	2
Number of neurons in the first fully connected layer	10
Number of neurons in the second fully connected layer	5
Number of neurons in the third fully connected layer	2
Number of epochs	15
Number of minimal batches	256