Fig. 9.

The accuracy and losses against the number of epochs of the proposed model without 1 × 1 convolution layer, where a training and validation for each fold, b training accuracy for different fold, and c an accuracy and losses for fold 5