An information-maximization approach to blind separation and blind deconvolution
暂无分享,去创建一个
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 distributions...