摘要
深度学习是近年来学界、业界一个十分热门的话题。在短短几年时间里,深度学习颠覆了语音图像识别、文本理解等众多领域,极大程度上推进了人工智能的快速发展。本文尝试从深度学习的定义与发展历程出发,阐释深度学习的基本原理,并将深度学习与传统的神经网络进行对比。进而,对深度学习的几种常见方法进行系统性归类阐述,并对接了行业研究的最新现状。文章的最后,对深度学习在管理实践当中的应用及深度学习带来的变革与挑战进行了发散性讨论。
Deep learning is a hot topic in no only academic but also business area in recent years, deep learning has overturned many areas such as speech image recognition and text understanding, and has greatly promoted the rapid development of artificial intelligence, in just a few days. This paper begins with the definition and development of deep learning, explains the basic principles of deep learning, and compares deep learning with traditional neural networks. Further, this paper systematically categorizes several common methods of deep learning and correlates the latest tensions in industry research. At the end of the paper, a divergent discussion was held about the application of deep learning in management practices and the changes and challenges brought about by deep learning.
作者
肖泽中
Xiao Zezhong(Renmin University of China,Beijing 100872,China)
出处
《未来与发展》
2018年第9期57-63,共7页
Future and Development
关键词
深度学习
神经网络
分类
训练
deep learning
neural network
classification
training