Title: Bayesian spectral analysis of replicated time series
Authors: Zeda Li - City University of New York (United States) [presenting]
Abstract: Technological advances have facilitated an explosion in the number of studies that collect time series data from multiple subjects to better understand how power spectra are associated with cross-sectional covariates. However, analyzing such data poses significant challenges due to the complicated structure of the power spectrum and the dynamic dependence structure between power spectra. While methods for single time series are rather extensive, existing methods for estimating the time-varying spectrum of a replicated time series are relatively few. We will introduce a flexible spectral analysis framework for replicated time series and explore open research questions in replicated spectral analysis brought about by complicated modern data structures.