Literature Review of Deep Learning Research

Abstract

t
Deep learning is one of most research fields of artificial
intelligence, which has been widely used from high-tech
companies in the worlds such as Google and other high-tech
companies have increased their exploitation in all fields of
artificial intelligence. Deep learning is significant advance in
neural networks to solve must of problem-relevant has big
features, it has applied in the speech processing, online
advertising, natural language processing, image recognition and
so on. This paper summarizes the latest advanced in deep
learning. First, review three basic types of deep learning,
including multilayer perceptron’s, convolutional neural
networks, and recurrent neural networks. On this basis, further
analysis of the emerging new convolutional neural network and
recurrent neural network. Then this paper summarizes the deep
learning in many fields of artificial intelligence, including speech
processing, computer vision and natural language processing.
Finally, discussed current problems of deep learning and given
the corresponding possible solutions.