ANNs are trained using the backpropagation of error algorithm, which involves calculating model-generated errors on the training dataset and updating the model weights to fix those errors. The limitation of this method of training is that the instances from each class are treated the same, which ...