Research Article

Improved Feature Learning: A Maximum-Average-Out Deep Neural Network for the Game Go

Figure 4

MAO network.