Slope Stability Assessment Based on Finite Element Method and XGBoost

Abstract Lam Dong, a mountainous province in Vietnam’s Central Highlands, is characterized by high elevations and substantial rainfall, leading to an increased risk of landslides. Assessing slope stability in this region is therefore essential. This study investigates the effects of rainfall discharge and geotechnical parameters on the Factor of Safety (FoS) of slopes. Using the Finite Element Method (FEM) in PLAXIS 2D, 500 simulations were conducted with four input variables: rainfall discharge (q), hydraulic conductivity (k), cohesion (c), and internal friction angle (φ). The simulation results were analyzed using the Extreme Gradient Boosting (XGBoost) machine learning algorithm to determine the relative influence of each parameter on slope stability. The findings show that geotechnical and hydrogeological conditions affect slope stability to varying degrees, with the internal friction angle (φ) identified as the most influential factor. These results offer valuable insights into slope behavior and support more effective landslide prediction and risk management strategies in the region.