Prediction of Lateral Load Capacity of Pile in Clay Using Multivariate Adaptive Regression Spline and Functional Network

This paper discusses the use of multivariate adaptive regression splines (MARS) and functional networks (FN) for prediction of the lateral load capacity of piles in clay. The results obtained from MARS and FN have been compared with different empirical models and artificial neural network in terms of statistical parameters such as correlation coefficient (R), Nash–Sutcliff coefficient of efficiency (E), absolute average error, maximum average error and root mean square error. Based on the statistical parameters, MARS and FN were found to have a better predictive capacity. Predictive equations are provided based on the MARS and FN model. A sensitivity analysis is also presented to determine the importance of inputs in prediction of the lateral load capacity of piles.