We are proud to invite back another former New Yorker, Ivor Cribben, for this month's meetup. Given the nature of Dr. Cribben's talk we will be holding this meetup at NYU's medical center. Yes, it is all the way by the East River, but it will be a good venue for this material.
Ivor Cribben is an Assistant Professor of Statistics in the Department of Finance and Statistical Analysis at the Alberta School of Business, University of Alberta. He holds a BA in Mathematics from Trinity College Dublin, an MSc in Applied Statistics from University College Dublin, and a PhD in Statistics from Columbia University.
Recently in brain imaging studies there has been an increased interest in quantifying changes in connectivity between brain regions over the experimental time course to provide a deeper insight into the fundamental properties of brain networks. The application of network science and graphical modelling has been instrumental in these analyses and enabled the examination of the brain as an integrated system. In this talk, I will outline some novel data-driven statistical methods that estimate the dynamic connectivity or networks between pre-defined brain regions using ideas from graphical models and time series analysis. I will also address the challenge of estimating a group-level dynamic connectivity structure across subjects. Finally, I will discuss the application of these new methods to simulated vector autoregression (VAR) data, to single-subject and multi-subject functional magnetic resonance imaging (fMRI) data from task-related and resting-state experiments as well as to Electroencephalography (EEG) data from a subject with epilepsy. Techniques that will be discussed include: working with fMRI and EEG data, graphical lasso, stationary bootstrap, and Granger Causality. R packages used include: glasso, gplot and JGL.
Pizza starts at 6:30, the presentation begins at 7 and then we will head to a bar.
Please note the new location and allow extra time to find the room.