The Counterpropagation Network (CPN), introduced by Robert Hecht-Nielsen in 1987, is a hybrid feedforward neural network that combines unsupervised competitive learning with supervised outstar learning. Unlike pure backpropagation networks, CPN trains in two