A0499
Title: Covariate adjusted mixed membership models for functional data
Authors: Donatello Telesca - UCLA (United States) [presenting]
Abstract: Mixed membership modelling in the context of functional data analysis is discussed. The aim is to propose to leverage the multivariate KL representation of a stochastic process to induce a probabilistic representation of mixed membership to pure membership processes. In this context, covariate adjustment is discussed about both the mean and covariance functions. The motivation comes from applications in functional brain imaging through electroencephalography.