Machine Learning-Based Prediction of Soil Bearing Capacity in Geotechnical Engineering Applications

Abstract

Accurate prediction of soil bearing capacity is essential in geotechnical engineering to ensure the safety, stability, and cost-effectiveness of civil infrastructure projects. Traditional empirical and analytical approaches often face limitations in handling complex soil behavior and nonlinear relationships among geotechnical parameters. This study presents a machine learning-based framework for predicting soil bearing capacity in geotechnical engineering applications to improve prediction accuracy and support intelligent foundation design. The proposed methodology integrates geotechnical datasets, statistical analysis, and machine learning algorithms to model the relationship between soil properties and bearing capacity performance. Parameters such as soil density, moisture content, cohesion, angle of internal friction, and foundation dimensions were utilized as input variables for predictive analysis. Multiple machine learning techniques, including artificial neural networks, support vector machines, and decision tree-based models, were employed to evaluate prediction efficiency under varying soil conditions. Experimental and computational analyses were conducted using field and laboratory data to validate the developed predictive models. The results demonstrated that machine learning algorithms significantly improved prediction accuracy, reduced computational complexity, and enhanced generalization capability compared with conventional empirical methods. The developed models also exhibited strong adaptability in handling nonlinear soil behavior and heterogeneous geotechnical conditions. Furthermore, feature analysis identified the most influential soil parameters affecting bearing capacity estimation, supporting optimized foundation planning and risk assessment. The study concludes that machine learning-based predictive techniques provide an effective and reliable solution for soil bearing capacity estimation in modern geotechnical engineering applications.