Under the background of open sourcing big database, in order to analyze and predict the ground surface subsidence caused by shield tunneling in the tunnel construction, and to accommodate more factors that affect the ground surface settlement and improve the accuracy of settlement prediction, this paper summarized the modeling principles of support vector machines. Based on this, the support vector machine (SVM) was applied to the ground surface settlement prediction in this paper. A case study was carried out on the South Hongmei Road Tunnel construction with eight factors selected as input features including soil parameters, shield parameters, tunnel depth, etc., and the ultimate ground surface settlement was selected as the output target value. Cross validation was used to determine the optimal parameters of LIB-SVM and afterwards the prediction model was established. The ground surface settlement caused by shield tunnel construction was predicted and compared with the in situ measurement. The results showed that the data measurement is almost reproduced by the prediction with an error within 5%. The outcome of this research indicates that the SVM method is feasible in practical prediction in the ground surface settlement caused by shield tunnel construction, which offers a new approach on the research of the tunnel engineering.