Apply the learning for maximum effect