Abstract:
|
Motivated by applications in biomedicine, we review several ideas associated with modeling functional data. From a Bayesian perspective, flexible probability models are formulated and related to classical contributions in the theory of Gaussian processes and rank-regularized estimation. This basic construction is extended to represent highly-structured observations in functional brain imaging. Specifically, we discuss applications to electroencephalography (EEG) data, collected in connection with studies of neurocognitive development in children with Autism Spectrum Disorder (ASD).
|