A0244
Title: Clustering spatial functional data using a geographically weighted Dirichlet process
Authors: Guanyu Hu - The University of Texas Health Science Center at Houston (United States) [presenting]
Abstract: A Bayesian nonparametric clustering approach is proposed to study the spatial heterogeneity effect for functional data observed at spatially correlated locations. A geographically weighted Chinese restaurant process equipped with a conditional autoregressive is considered prior to fully capturing the spatial correlation of function curves. To sample efficiently from the model, a prior called Quadratic Gamma is customized to ensure conjugacy. A Markov Chain Monte Carlo algorithm is designed to infer simultaneously the posterior distributions of the number of groups and the grouping configurations. The superior numerical performance of the proposed method over competing methods is demonstrated using simulated examples and various applications.
