Abstract: We derive a new self-organizing learning algorithm that maximizes the information transferred in a network of nonlinear units. The algorithm does not assume any knowledge of the input ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results