The objective of this study is to develop a predictive model for ground vertical displacement around deep excavation sites by combining the Finite Element Analysis (FEA) method with Machine Learning techniques. The research focuses on assessing ground displacement under the influence of complex geotechnical factors, particularly in mixed clay conditions and variations during the excavation process. The approach involves using FEA simulations to model the excavation in different stages, then using this data to train Machine Learning models, such as Random Forest, to predict displacement at various locations. The combination of FEA and Machine Learning not only optimizes the prediction process but also reduces computation time compared to traditional methods. The results indicate that the predictive model has high accuracy, with performance metrics such as MSE, RMSE, and R2 yielding very good values, demonstrating the model's ability to accurately predict displacement changes at different locations within the excavation site. The model has proven effective in minimizing errors and enhancing predictive capabilities, thus aiding better management of the construction process. The integration of FEA and Machine Learning is an effective approach for improving the prediction and management of ground vertical displacement, contributing to enhanced safety and optimization in deep excavation projects and urban construction.