Multivariate adaptive regression splines model for reinforced soil foundations

Abstract

In this study, a multivariate adaptive regression splines (MARS) model has been developed to predict the settlement of shallow reinforced sandy soil foundations (RSSFs). The potential of the MARS model is validated comparatively with four other robust artificial intelligence/machine learning regression models, namely extreme learning machines (ELM), support vector regression (SVR), Gaussian process regression (GPR), and stochastic gradient boosting trees (SGBT). The pertinent data retrieved from previously published well-established scientific studies have been used to calibrate and validate the data-driven intelligent machine learning models. The predictive strength of all the modelling tools mentioned above were assessed via several statistical indices. Moreover, the predictive ability and reliability of the developed models were also corroborated with ranking criteria and external validation analysis. The results as obtained have shown that the MARS modelling technique attains the superior veracity in predicting the settlement of reinforced foundations.