A0209
Title: A generalized estimating equation approach to multivariate adaptive regression splines
Authors: Jakub Stoklosa - University of New South Wales (Australia) [presenting]
David Warton - University of New South Wales (Australia)
Abstract: Multivariate adaptive regression splines (MARS) is a popular nonparametric regression tool often used for prediction and uncovering important data patterns between the response and predictor variables. The standard MARS algorithm assumes normality and independence between continuous response variables. We extend MARS to generalized estimating equations, and we refer to this MARS-for-GEEs algorithm as MARGE. Through simulation we show that the proposed algorithm has improved predictive performance compared with the original MARS algorithm when using correlated and/or non-normal response data and is competitive with alternatives in the literature, especially for problems with multiple interacting predictors. The proposed algorithm is applied to various ecological data types.
