A1533
Title: Coarse-to-fine spatial GLMM for scalable prediction and multiscale analysis
Authors: Daisuke Murakami - The Institute of Statistical Mathematics (Japan) [presenting]
Alexis Comber - University of Leeds (United Kingdom)
Takahiro Yoshida - The University of Tokyo (Japan)
Narumasa Tsutsumida - Fujitsu Limited (Japan)
Chris Brunsdon - Maynooth University (Ireland)
Tomoki Nakaya - Tohoku University (Japan)
Abstract: Although a recent study suggested that coarse-to-fine learning provides a fast and flexible framework for large-scale spatial process modeling, the method was originally developed for Gaussian responses, limiting its applicability. To address this limitation, the coarse-to-fine spatial modeling framework was extended to accommodate spatial generalized linear mixed models (GLMMs), with a particular focus on count and binary data. The resulting spatial GLMMs efficiently addresses the degeneracy problem often encountered in conventional spatial GLMMs. The performance of the proposed spatial GLMMs was evaluated in terms of spatial prediction and multiscale feature extraction via Monte Carlo experiments and an empirical application. The developed method is implemented in an R package spCF.
