Abstract: With the continuous application of computer technology to the industrial internet of things, data transmission in industrial systems needs to support high real-time, high reliability, high bandwidth and massive connections. Traditional networks can no longer meet these needs, and 5G networks have become a research hotspot in the field of industrial internet of things due to their superior performance such as high speed, low latency, support for massive connections and good mobility. We review the 5G machine learning methods for industrial networks, first analyzes the key technologies such as large-scale antennas, terminal direct connection, mobile edge computing and heterogeneous ultra-dense networking in the field of 5G network communication technology, then introduce artificial intelligence technology and machine learning technology as an important part of it, summarize and prospect the methods of introducing machine learning technology into 5G network to solve specific problems, and finally put forward the future research trend of 5G communication technology.