ECG Heartbeat Classification Based on an Improved ResNet-18 Model

Algorithm 1

The training algorithm of the improved ResNet-18 model.

Data: preprocessed ECG signals

Result: classified heartbeats

Begin:

 1. Initialize the layer learning rate α, total number of max iteration epoch, min error , Total Batch;

 2. Find the value ;

 3. Generate random weights of the ResNet-18;

 4. ResNet-18 model = Init ResNet-18 model();

 5. while and do:

 6.  Initialize ;

 7.  for to Total Batch training do:

 8.   ;

 9.   Update ;

 10.   ;

 11.  end for.

 12.  Iteration ;

 13. end while.

Return: Output with minimum .