Vector generalized linear and additive extreme value models (Q928489)

A methodology of statistical analysis of extreme data on the base of vector generalized linear (VGLM) and vector generalized additive (VGAM) models by the R software is discussed. Procedures of the VGAM package are applied to annual maximum sea levels data, rainfall data and to a simulation example. The authors describe briefly the VGLM models, the iteratively reweighted least squares algorithm for fitting Euclidian parameters and the vector splines penalized likelihood technique for fitting the additive nonparametric components of the model.