Theoretical Foundations for Specific Architectures
摘要
In the previous chapters, we have presented an overview of the generalizability of neural networks. This chapter discusses the influence of some specific network structures. Convolutional neural networks (CNNs) introduce convolutional layers into deep learning, which have been widely applied in computer vision, natural language processing, and deep reinforcement learning. Recurrent neural networks (RNNs) possess a recurrent structure and shows its promising performance in the processing and analysis of sequential data.