分享

NoProp: Training Neural Networks without Back-propagation or Forward-propagation

212
查看论文
热度