A0520
Title: Spatially-varying coefficient models with structure identification
Authors: Guannan Wang - College of William & Mary (United States) [presenting]
Zhiling Gu - Iowa State University (United States)
Xinyi Li - Clemson University (United States)
Lily Wang - Iowa State University (United States)
Abstract: A general framework of spatially varying coefficient models with structure identification is introduced, which involves automatic detection of the significance and types of effects (spatially varying or constant) of various factors on the response of interest. The proposed method can efficiently identify spatially varying coefficient components, enhancing computational efficiency and statistical power for downstream analysis. To provide a solid theoretical foundation for the proposed method, the consistency and asymptotic normality of the constant estimators are rigorously established. Moreover, extensive Monte Carlo simulations are conducted to examine the efficacy of the proposed method in identifying the true model structure and improving estimation and prediction accuracy. Additionally, the practical utility of the framework is illustrated through an analysis of particulate matter (PM), which provides valuable insights into the influence of environmental factors on observed PM.
