Effective stress parameter in unsaturated soils: an evolutionary-based prediction model

Abstract

Deformations and failures in unsaturated soils are influenced directly by the effective stress calculated using the stress equation affected by the effective stress parameter. A data-mining-based approach was implemented in this research to develop a prediction model for the effective stress parameter in unsaturated soils. The proposed modelling approach took an evolutionary computing technique to find polynomial models that are structured and explicit. A combination of the genetic algorithm and the least squares approach was implemented to search for the most suitable polynomial structures and the corresponding term parameters in the models. A set of unsaturated soil triaxial test results were used to develop the model. The model was evaluated based on its performance for making predictions using unseen data to validate generalisation capabilities. Model predictions were compared to laboratory and artificial neural-network-based model performance data. A sensitivity analysis assessed the level and form of contributions that input parameters made to the developed model. The developed model accurately captured and redeveloped the intrinsic connections between the introduced input parameters and produced effective stress parameter predictions which competed with the artificial neural network model in terms of accuracy of predictions and outperformed it with regards to the model structure, simplicity and transparency.