Simplified Information Maximization for Improving Generalization Performance in Multilayered Neural Networks

Figure 13

From an initial state to a maximum information state.