DONDENA Seminar - Donatello Telesca
You may follow the seminar at the following link.
“Unlocking heterogeneity in the developing brain: a Bayesian view of electroencephalography”
SPEAKER: Donatello Telesca (UCLA)
ABSTRACT:
The human brain and its functional relation to biobehavioral processes like motor coordination, memory formation and perception as well as pathological conditions like Parkinson’s disease, epilepsy and Autism have been a subject of intense scientific scrutiny. Our work is motivated by neurodevelopmental studies of children and adolescents, with a specific focus on implicit learning and resting state brain function. We are especially interested in characterizing brain function patterns and heterogeneity in Autistic children, where clinical presentation is highly variable, with heterogeneity in relation to medical conditions, behavioral challenges, and degree of intellectual impairments. In this context, we consider statistical analysis in elecroencephalography (EEG), a well-established noninvasive method for measuring spontaneous and event-related electrical activity across brain regions. This lecture aims to highlight a Bayesian view of statistical inference for EEG data, with the goal of coupling flexible semiparametric modeling with a formal probabilistic characterization of complex observational processes. Special consideration will be given to Bayesian unsupervised and semisupervised learning for multilevel data, with special emphasis on Bayesian functional data analysis. Finally, we show how the careful application of hierarchical modeling tools allow for the a formal characterization of fundamental neuroscientific concepts, including a formal quantification of the "Autism spectrum".
BIO:
After graduating with a laurea in Economics from Bocconi, Dr. Telesca received a Ph.D. in Statistics from the University of Washington. His research interests include Bayesian methods in multivariate statistics, functional data analysis, statistical methods in bio- and nano-informatics. His interest in highly structured multivariate observations extends to the field of nano-informatics, where he worked on the formulation of Statistical approaches to Quantitative Structure Activity Relationships models. His work in functional data analysis has focused on stochastic models targeting the timing of latent events through time-warping and has seen applications ranging from bio-informatics to criminology. Dr. Telesca is a fellow of the American Statistical Association, a member of the California NanoSystems Institute and the UCLA Jonsson Comprehensive Cancer Center.