photo 3

Research areas:

  • Time series analysis
  • Bayesian models
  • Statistical models for brain images and signals

UCI/UCSC Space-Time Research Group

Current research projects:

NSF SES-1461497 Collaborative Research: Bayesian State-Space Models for Behavioral Time Series Data. This is a collaborative project with Prof. Hernando Ombao (Statistics, UCI).

The focus of this project is on developing statistical methods for studying temporal data derived from functional magnetic resonance imagining (fMRI) and local field potentials (LFP). These methods also are applicable to other types of brain signals, such as electroencephalograms (EEG) and magnetoencephalograms (MEG). The project develops Bayesian state-space models for activation and connectivity in fMRI data. These models simultaneously estimate the hemodynamic behavior in local areas of the brain and the inter-dependence between brain regions in a network, while taking into account variations across subjects and differences across experimental conditions. The Bayesian state-space models and related inferential tools are extended to consider associations between the neural-derived brain signals and behavioral data under the context of behavioral experiments.



Students currently affiliated with the project:

  • Cheng-Han Yu (PhD Candidate UCSC). Cheng-Han is working with complex-valued fMRI data.

Students and researchers previously affiliated with the project:

  • Jing Cao (MS 2015, University of California Santa Cruz). Jing currently works for The Climate Coorporation, San Francisco, CA.
  • Zhe Yu (PhD 2015, University of California Irvine). Zhe currently works for Apple, Cupertino, CA.


NSF DMS-1407838 Bayesian nonparametric methods for spectral analysis of complex brain signals. This is a collaborative project with Prof. Athanasios Kottas (AMS, UCSC) 

This project develops Bayesian nonparametric methods and computational tools for frequency-domain analysis of multiple time series with application to the analysis of brain signals. Novel and flexible mixture models are used to represent the spectral characteristics of multiple time series. The methods being developed have the following key features: (i) they provide flexible representations of the spectral densities of multiple signals as well as computational feasibility (ii) they allow researchers to investigate clustering patterns of multiple time series with similar spectral characteristics, and (iii) they incorporate hierarchical settings that can appropriately accommodate neuroscience data sets involving multiple trials, multiple subjects and/or relevant covariates.


Students currently affiliated with the project:

  • Annalisa Cadonna  (PhD Candidate, AMS UCSC). Annalisa received the People’s Choice Award in the 2015 UCSC Grad Slam Competition for her video Flexible Spectral Modeling for the Analysis of Complex Brain Signals. Congratulations Annalisa!!!


Additional very recent projects/collaborators:

  • With Prof. Rajarshi Guhaniyogi (AMS, UCSC) and AMS PhD student Daniel Spencer, we have just recently started a research collaboration to develop Bayesian tensor regression approaches for neuroimaging data. 
  • We have been working with Prof. Daniel Rowe (Marquette University) in the analysis of complex-valued fMRI signals.