Sequential Monte Carlo and Protein Structure/Funding Opportunities for Statisticians at NSF

GUEST LECTURER
Rong Chen

AFFILIATION
University of Illinois at Chicago and NSF

ABSTRACT
Part I


Sequential Monte Carlo and Protein Structure

Rong Chen
Department of Information and Decision Sciences
University of Illinois at Chicago



The sequential Monte Carlo (SMC) methodology emerged in the fields of statistics and engineering has shown a great promise in solving a large class of highly complex inference and optimization problems. It is a family of Monte Carlo techniques that is based on the 'growth principle' in dealing with high dimensional complex systems. Specially, a high dimensional problem is decomposed into a sequence of low dimensional problems which forms a stochastic dynamic system. Weighted Monte Carlo samples are drawn from each of simpler component, and their combination then forms a Monte Carlo sample of the original system.

In this talk, we present SMC approaches in studying geometric properties of proteins. Specifically we introduce the general strategy of intermediate distribution design, look-ahead, multi-grid sampling and constraint enforcement. It is used to generate protein structures with certain geometric constraints, under a discrete off-lattice model.



Part II

Funding Opportunities for Statisticians at NSF

Rong Chen
National Science Foundation

In this talk, some of the funding opportunities for statisticians at NSF will be introduced, including the regular statistics program and programs in the mathematical priority area. Some tips on proposal writing will be offered. Suggestions, comments, and discussion on future direction of statistics are welcome.

DATE & TIME
Thursday, March 16, 2006 -- 9:30 AM

DURATION
1 hour

LOCATION
Executive Room #228

CONTACT PERSON
Ming Yuan