WorkshopsTime-varying Additive Model for Longitudinal Data
81
reads
Additive model is an effective dimension reduction approach that provides flexibility to model the relation between a response variable and key covariates. The literature is largely developed for scalar response and vector covariates. In this paper, more complex data is of interest, where both the response and covariates may be functions. A functional additive model is proposed together with a new smooth backfitting algorithm to estimate the unknown regression functions, whose components are time-dependent additive functions of the covariates. Due to the sampling plan, such functional data may not be completely observed as measurements may only be collected intermittently at discrete time points. We develop a unified platform and efficient approach that can cover both dense and sparse functional data and the needed theory for statistical inference. The oracle properties of the component functions are also established. Title: Dimension Reduction for Functional Data Abstract: Functional data are intrinsically infinite dimensional, so dimension reduction methods are often needed to handle it. The approaches might differ for dense and sparse functional data. In this talk, we'll review some of the approaches, including principal component based methods and dimension reduction regression models.
reads
Jane Ling Wang
2012-12-21
14:20:00 - 15:10:00
308 , Mathematics Research Center Building (ori. New Math. Bldg.)
Additive model is an effective dimension reduction approach that provides flexibility to model the relation between a response variable and key covariates. The literature is largely developed for scalar response and vector covariates. In this paper, more complex data is of interest, where both the response and covariates may be functions. A functional additive model is proposed together with a new smooth backfitting algorithm to estimate the unknown regression functions, whose components are time-dependent additive functions of the covariates. Due to the sampling plan, such functional data may not be completely observed as measurements may only be collected intermittently at discrete time points. We develop a unified platform and efficient approach that can cover both dense and sparse functional data and the needed theory for statistical inference. The oracle properties of the component functions are also established. Title: Dimension Reduction for Functional Data Abstract: Functional data are intrinsically infinite dimensional, so dimension reduction methods are often needed to handle it. The approaches might differ for dense and sparse functional data. In this talk, we'll review some of the approaches, including principal component based methods and dimension reduction regression models.