An area-specific stick breaking process for spatial data (Q2633423)

The authors introduce a new Bayesian nonparametric model for spatial areal data which allows for unknown neighborhood structures. This model is based on an area-specific stick breaking process for distributions of random effects with the spatially-dependent weights arising from the block averaging of underlying continuous surfaces. The authors illustrate the flexibility of this new prior in capturing heterogeneity in spatial dependency across areas and investigate the methodology numerically with a real dataset.