Constrained Sequential Monte Carlo
Rong
Chen

Rutgers University

Thursday, April 8, 2010 - 3:30pm

The sequential Monte Carlo (SMC) methodology has shown a great promise in solving a large class of highly complex inference and optimization problems. Although it was originally designed to solve on-line filtering and smoothing of non-linear non-Gaussian state space models, it has been shown to be equally powerful in dealing with fixed-dimensional problems, utilizing a sequential decomposition principle. In this talk we discuss issues and efficient implementations of SMC for dealing with high dimensional distributions that are defined on restricted and ill-shaped spaces. Examples in bioinformatics (RNA and Protein geometric structures) and financial engineering (generating diffusion bridges) are presented.