Title: A three-step nonparametric regression approach for analyzing spatio-temporal data
Authors: Peihua Qiu - University of Florida (United States) [presenting]
Abstract: Spatio-temporal data are common in practice. Research in proper analysis of spatio-temporal data has attracted much attention in different research communities, including statistics, epidemiology, geography, oceanography, environmental science, and more. Existing methods in the literature often employ parametric modelling with different sets of model assumptions. However, spatio-temporal data in practice usually have complicated structures, including complex spatial and temporal data variation, spatio-temporal data correlation, and data distribution. Because such data structures reflect the complicated impact of confounding variables, such as weather, demographic variables, life styles, and other cultural and environmental factors, they are usually too complicated to be described well by parametric models. We discuss a general modelling framework for analyzing spatio-temporal data based on nonparametric spatio-temporal regression. The suggested model and its estimation can well accommodate the complicated structure of real spatio-temporal data described above. Both theoretical arguments and numerical studies show that our proposed method could work well in practice.