In order to improve the accuracy and efficiency for the inverse of physical and mechanical parameters of high arch dams, Jaya algorithm and Gaussian process machine learning theory were introduced into the field of dam safety monitoring, and an arch dam parameter inverse analysis method based on the Jaya-Gaussian process regression surrogate model was proposed. The Gaussian process regression surrogate model was used instead of the traditional finite element calculation, and three intelligent optimization algorithms were used to optimize the parameters. The results show that compared with PSO and GWO algorithm, Jaya algorithm not only has high inversion accuracy, fast convergence speed, and good stability, but also has good stability, and the proposed inverse analysis strategy saves more than 80% time compared with the inverse analysis method that directly calls finite element calculation. This method can not only meet the requirements of calculation accuracy, but also greatly reduce the calculation time, providing an efficient method for the inverse analysis of physical and mechanical parameters for high arch dams. Keywords: high arch dam; displacement inverse analysis; Gaussian process regression; surrogate model; Jaya algorithm 〖FL