BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20250514T112735EDT-9611vpZONV@132.216.98.100
DTSTAMP:20250514T152735Z
DESCRIPTION:A simple new approach to variable selection in regression\, wit
 h application to genetic fine-mapping\n\nMatthew Stephens\, University of 
 Chicago\n	Tuesday September 28\, 12-1pm\n	Zoom Link: https://mcgill.zoom.us/
 j/85428056343\n\nAbstract: We introduce a simple new approach to variable 
 selection in linear regression\, and to quantifying uncertainty in selecte
 d variables. The approach is based on a new model – the ``Sum of Single Ef
 fects'' (SuSiE) model -- which comes from writing the sparse vector of reg
 ression coefficients as a sum of ``single-effect'' vectors\, each with one
  non-zero element. We also introduce a corresponding new fitting procedure
  -- Iterative Bayesian Stepwise Selection (IBSS) -- which is a Bayesian an
 alogue of stepwise selection methods. IBSS shares the computational simpli
 city and speed of traditional stepwise methods\, but instead of selecting 
 a single variable at each step\, IBSS computes a {\it distribution} on var
 iables that captures uncertainty in which variable to select. The method l
 eads to a convenient\, novel\, way to summarize uncertainty in variable se
 lection\, and provides a Credible Set for each selected variable. Our meth
 ods are particularly well suited to settings where variables are highly co
 rrelated and true effects are sparse\, both of which are characteristics o
 f genetic fine-mapping applications. We demonstrate through numerical expe
 riments that our methods outperform existing methods for this task.\n
DTSTART:20210928T160000Z
DTEND:20210928T170000Z
LOCATION:CA\, QC
SUMMARY:QLS Seminar Series - Matthew Stephens
URL:/qls/channels/event/qls-seminar-series-matthew-ste
 phens-333115
END:VEVENT
END:VCALENDAR