Title: A robust version of GAMLSS
Authors: William Aeberhard - Stevens Institute of Technology (United States)
Eva Cantoni - University of Geneva (Switzerland)
Giampiero Marra - University College London (United Kingdom)
Rosalba Radice - Cass Business School (United Kingdom) [presenting]
Abstract: A robust version of generalised additive models for location, scale and shape is discussed where any parameter of the distribution can be specified as function of additive predictors allowing for several types of covariate effects (e.g., linear, non-linear, random and spatial effects). The estimation approach permits all models parameters to be estimated robustly by limiting the influence of deviating data points on each log-likelihood contribution. We evaluate the empirical performance of the proposed method through simulation experiments. We also illustrate the use of this approach on functional magnetic resonance imaging measurements for a human brain subject to a particular experimental stimulus.