|
Abstract:
|
Generalized additive models (GAMs) provide flexible models for a wide array of data sources. In the past, improvements of GAM estimation have focused on the smoothers used in the local scoring algorithm used for estimation, but poor prediction for non-Gaussian data motivates the need for robust estimation of GAMs. In this presentation, rank-based estimation, as a robust and efficient alternative to the likelihood-based estimation, of GAMs is proposed. It is shown that rank GAM estimators can be obtained through iteratively reweighted likelihood-based GAM estimation which we call the iteratively reweighted rank quasi-likelihood (IRRQL). Simulation experiments support the use of rank-based GAM estimation for heavy-tailed or contaminated sources of data. Application of rank GAM estimation via IRRQL on a fisheries dataset, a field which commonly uses GAMs for their high degree of flexibility in modeling complex systems, demonstrates that the proposed method provides better fit and often better prediction in comparison to the classical likelihood GAM fits.
|
