I guess they may depend on the size of training data (to avoid overfitting) and on the efficiency desired.
Thank you, Riccardo
I guess they may depend on the size of training data (to avoid overfitting) and on the efficiency desired.
Thank you, Riccardo