IMS
Session Slot: 2:00- 3:50 Wednesday
Estimated Audience Size:
AudioVisual Request:
Session Title: Special Invited Papers II
Theme Session: No
Applied Session: No
Session Organizer: Lindsay, Bruce The Pennsylvania State University
Address: 422 Thomas Building, Department of Statistics, University Park, PA 16802
Phone: (814) 865-1220
Fax: (814) 863-7114
Email: bgl@psu.edu
Session Timing: 110 minutes total (Sorry about format):
Session Chair: Lindsay, Bruce The Pennsylvania State University
Address: 422 Thomas Building, Department of Statistics, University Park, PA 16802
Phone: (814) 865-1220
Fax: (814) 863-7114
Email: bgl@psu.edu
1. Inference in High Dimension Models
Murphy, Susan, University of Michigan
Address: The University of Michigan 1440 Mason Hall Ann Arbor, MI 48109-1027
Phone: (313) 763-3519
Fax: (313) 763-4676
Email: murphy@stat.psu.edu
van der Vaart, A.W., Free University, Amsterdam
Abstract: Likelihood inference in high dimensional models appears, at best, to be characterized by variety: variety in the form of the likelihood, variety in distributional properties, variety in rates of convergence, variety in the norms, etc. Even in apparently ``regular/smooth'' models, there does not appear to be any generally applicable approaches for ascertaining asymptotic properties. This state of affairs is reminiscent of maximum likelihood estimation in smooth parametric models during the pre-Cramér days.This talk will discuss examples illustrating the above variety, pointing out areas in much need of research. In models in which a finite dimensional parameter is of primary interest, it is possible to give a general methodology for ascertaining asymptotic properties of inferences based on a maximum likelihood estimator. This methodology assumes smoothness similar to Cramér's classical conditions and can be applied to a wide variety of models which are not smooth in the nuisance parameter. Aspects which need further development will be discussed.
List of speakers who are nonmembers: