Vector Generalized Linear and Additive Models

Abstract

Vector Generalized Linear and Additive Models Thomas William Yee Department of Statistics, University of Auckland, New Zealand ? This talk will give an overview of VGLMs and VGAMs. We will look at the unified statistical framework and a software implementation of it (the R package VGAM). The class of VGLMs/VGAMs is very large and contains many statistical models, e.g., univariate and multivariate distributions, categorical data analysis, extreme value analysis, correlated binary data and nonlinear regression. The topics covered include the the IRLS algorithm, the vector smoothing problem (especially vector cubic smoothing splines), constraint matrices, reduced-rank VGLMs, and species-site ordination (in ecology).