Novel Hybrid XGBoost Model to Forecast Soil Shear Strength Based on Some Soil Index Tests

When building geotechnical constructions like retaining walls and dams is of interest, one the most important factors to consider soil’s shear strength parameters. This study makes an effort propose a novel predictive model strength. The implements extreme gradient boosting (XGBoost) technique coupled with powerful optimization algorithm, salp swarm algorithm (SSA), predict various soils. To do this, database consisting 152 sets data prepared where (τ) soil considered as output some index tests (e.g., dry unit weight, water content, plasticity index) are set inputs. designed tuned using both effective parameters XGBoost SSA, accurate introduced in this study. prediction performance SSA-XGBoost assessed based on coefficient determination (R2) variance account for (VAF). Overall, obtained values R2 VAF (0.977 0.849) (97.714% 84.936%) training testing sets, respectively, confirm workability developed forecasting investigate generalization, tested another 30 (validation data). validation results 0.805) suggest proposed model. findings that when cannot be determined directly, hybrid XGBoost-SSA can utilized assess parameter.