第1课 Reducing the dimensionality of data with neural networks.
第2课 A fast learning algorithm for deep belief nets.
第3课 Densely Connected Convolutional Networks
第4课 A guide for convolution arithmetic for deep learning
第5课 Imagenet classification with deep convolutional neural networks.
第6课 Speech recognition with deep recurrent neural networks
第7课 Decoupled neural interfaces using synthetic gradients
第8课 Deep Learning without Poor Local Minima