Database, Model Parameters, Predictions, and Automated Calculator for the Effective Stress Parameter of Unsaturated Soils

This supplemental material supports the manuscript “Physics-Guided Sparse Gaussian RBF Neural Network for Closed-Form Prediction of the Effective Stress Parameter of Unsaturated Soils.” The workbook contains the compiled database, training-validation-testing partition, Gaussian RBF model parameters, record-level predictions, model-performance information, and an automated calculator for predicting Bishop’s effective stress parameter, χ, from user-specified hydraulic and mechanical inputs.