[PAST EVENT] Mathematics Colloquium and EXTREEMS-QED Lecture: Todd Ogden (Columbia University)
ABSTRACT: One goal in precision medicine is to make optimal patient-specific treatment decisions using available baseline data. One important application involves complex data in the treatment of neuropsychiatric disorders, including brain imaging or other functional data. Such data which can be incorporated into regression models of treatment outcome using functional data analytic techniques. Thus, obtaining useful and interpretable fits in such a situation generally involves some sort of dimension reduction or penalization, taking into account the particular (spatial) structure of the data. This talk will introduce the problems and describe some basic strategies that have proven effective in this context.