With the development of engineering information level and the monitoring technology in the field of shield tunnel, the recorded engineering data contains the internal information of tunneling equipment and its interaction with the external stratum. Machine learning has more application space than traditional modeling statistical analysis methods because of its strong data analysis ability and no requirement on prior theoretical formula and expert knowledge. Improving the efficiency and safety level of shield tunnel construction is helpful to deeply mine the collected information and data and analyze their internal relationship through machine learning method. This paper briefly describes the basic principle of machine learning methods, summarizes and analyzes its application in shield tunnel engineering. In particular, the progress on the equipment status analysis, shield performance prediction, geological parameters analysis, prediction of ground surface deformation and examination of tunnel hazard based on the machine learning method are summarized. Finally, the key problems to be solved so as to realize the intelligent shield tunnel engineering are analyzed and forecasted.