During the construction of a shield tunnel, it will disturb the surrounding ground and affect the use and structural safety of buildings around the tunnel. The geometric parameters of the tunnel, the operating parameters of the shield machine, and the geological parameters will affect the degree of disturbance. However, the existing theories and models are difficult to comprehensively consider the interaction of these factors, and it is difficult to accurately predict the response of the formation to solve the above problems. The research is based on the machine learning algorithm to establish a prediction model of stratum settlement caused by shield tunneling, which provides a new idea for real time prediction of the ground response caused by shield tunneling and risk reduction. The main results of this research are as follows: (1) propose a novel quantification method for geological parameters that can comprehensively consider the physical and mechanical properties of the rock,soil layers, and the geometric characteristics of depth and thickness and (2) establish a more robust proxy model and use the k-fold cross-validation method to enhance its performance.
