Comparative analysis of single and ensemble machine learning models in predicting soil shear strength

Mohamed Rabie

School of Computing and Engineering

Supervisors:

Dr Ibrahim Shaaban

School of Computing and Engineering

Dr Khaled Rabie

Qatar University

Soil shear strength is a critical factor in civil engineering, representing the soil’s ability to withstand shear stress. Accurately determining this property is essential for evaluating the stability of structures on or within the soil. While traditional methods are important, they can be complex and resource-intensive. As a result, advanced machine learning techniques have emerged as innovative solutions. This study examines the effectiveness of various ML algorithms, including Support Vector Regression and Gradient Tree Boosting, in predicting undrained shear strength of soil using a comprehensive dataset from prior research. The data was split into 80% for training and 20% for testing with 5-fold cross-validation to evaluate model performance based on statistical metrics such as MAPE, MAE, RMSE, and R^2. The results indicate that GTBR performed best with an R^2 value of 96.8%. This model has been seamlessly integrated into an online user-friendly interface to provide easy access for professionals seeking precise estimates of soil shear strength.