Statistics Seminar
| Date/Time: | Wednesday, 30 Oct 2013 from 4:10 pm to 5:00 pm |
|---|---|
| Location: | Snedecor 3105 |
| Cost: | Free |
| URL: | www.stat.iastate.edu |
| Contact: | Jeanette La Grange |
| Phone: | 515-294-3440 |
| Channel: | College of Liberal Arts and Sciences |
| Categories: | Lectures |
| Actions: | Download iCal/vCal | Email Reminder |
Fiducial inference was introduced by Fisher in 1930 in an attempt to overcome what he saw as a serious deficiency of the Bayesian approach to inference: the use of a prior distribution on model parameters even when no information was available. Fiducial inference created some controversy once Fisher's contemporaries realized that, unlike earlier simple applications involving only one single parameter, fiducial inference often led to procedures that were not exact in the frequentist sense and did not possess other properties claimed by Fisher.
In this talk we describe a modified version of Fisher's fiduical approach, termed generalized fiducial inference, and its applications to several modern problems, including ultrahigh dimensional regression and big data computations.
This is joint work with Jan Hannig and Randy C. S. Lai.