statistics colloquium

Event Type: 
Colloquium
Speaker: 
Richard Hahn
Event Date: 
Friday, January 29, 2016 -
3:00pm to 5:00pm
Location: 
SMLC 120
Audience: 
General PublicFaculty/StaffStudentsAlumni/Friends
Sponsor/s: 
Stat group

Event Description: 

Title: Bayesian regularized regression for treatment effect estimation 
  
Abstract:  
In this talk, I describe the use of regularization priors in the context of
treatment effect estimation where the number of control variables is large relative
to the number of observations.  A re-parametrized simultaneous regression model is
presented which permits prior specifications designed to overcome certain pitfalls
of the more conventional direct parametrization. The new approach is illustrated on
synthetic and empirical data. Finally, I describe the extension of our method to a
nonlinear mean regression model with homogeneous treatment effect. 

Event Contact

Contact Name: Yan Lu

Contact Phone: 505-277-2544

Contact Email: luyan@math.unm.edu