Prediction of Pile Bearing Capacity using Multivariate Adaptive Regression Spline (pp. 185-191)
Authors: (Pijush Samui, Centre for Disaster Mitigation and Management, VIT University, Vellore, India)
Abstract: This chapter examines the capability of Multivariate Adaptive Regression Spline(MARS) for prediction of bearing capacity(Q) of pile foundation. MARS is a procedure for adaptive nonparametric regression. It divides the dataspace in several, possibly overlapping regions and fitting truncated spline functions in each region. The data from a pile load test has been used to build the MARS model. MARS uses penetration depth ratio (l/d), mean normal stress (σm), and no of blows (n) as input parameters. The output of SVM model is Q. Equation has been developed for determination of Q of pile foundation. This chapter shows that the developed MARS model is a robust model for prediction of Q of pile foundation.