This study discusses various geotechnical parameters, such as total normal stress, effective suction, position, and displacement, that influence the active pore pressures. The analysis was incorporated with the method of the Finite Element Analysis with Machine Learning—the Random Forest model. Then, the aforementioned FEA method is used in simulating and extracting data for the different geotechnical factors on all other depth levels to have a complete understanding of how active pore pressures vary. The results obtained from FEA served as input for training the Random Forest model used in this study to predict active pore pressures with the main aim of assessing the importance of the observed variables. Theudos’ result of the study indicates that the Random Forest model is very efficient in predicting the active pore pressures with an R2 coefficient value equal to 0.999456. Total stress in the z-direction, Ϭzz, and total stress in the x-direction, Ϭxx, contributed to the modification in active pore pressures, amounting to 82.14% and 11.99%, respectively, while the rest, like effective suction and stresses in the y-direction, Ϭyy, are less influential. In this regard, the combination of FEA and Random Forest would allow not only fast and accurate active pore pressure prediction but also deeper insights into the relationships between geotechnical factors and active pore pressures, which could be used to optimize various geotechnical design and construction projects.