B1731
Title: Estimation and inference in generalized spatial partially linear varying coefficient models
Authors: Jingru Mu - Kansas State University (United States) [presenting]
Abstract: A class of generalized partially linear spatially varying coefficient models for data distributed over complex domains is considered. We approximate the varying coefficient functions and linear coefficients via penalized bivariate splines over triangulation that can handle the complex boundary of the spatial domain. The asymptotic normality of estimated constant coefficients and the consistency of estimated varying coefficients are built up under some regularity conditions. We further propose a model selection approach to identify covariates with constant and varying effects. The performance of the proposed method is evaluated by simulation studies and the crash data in Texas.
