Software by Kevin Murphy

http://www.cs.ubc.ca/~murphyk/Software/index.html

Kevin Murphy's publications

Kevin Murphy has done lot of work in dynamic bayesian networks. Here is a link of his publications.

http://www.cs.ubc.ca/~murphyk/papers.html#refereed

Sampling in dynamic bayesian networks

This is my term project report for STAT-330 (Statistical Inference I) class. The class was taught in Spring 2005 by Prof. Dipak Dey who is also Head of the Statistics Department at UConn.

Sampling in Dynamic Bayesian Networks
Abstract: In this paper we introduce Sequential Monte Carlo methods (SMC) for computing Bayesian inference in Dynamic Bayesian Networks (DBN). SMC methods provides an elegant stochastic approximation approach to compute inference in a large discrete DBN where inference requires to be computed at each time space and computations are exponential in the number of hidden nodes. We illustrate the particle filtering using a simple example. The paper discusses the particle filtering techniques such as likelihood weighting, Rao Blackwellized particle filtering which exploits the structure of DBNs to increase the efficiency of Bayesian inference. In the future work, we propose to extend some of the techniques discussed in this paper and apply them to solve a real world problem.

Project pdf file

Objective of this blog

I want to use this blog to share my knowledge about Bayesian networks. I will post links to key papers and articles related to Bayesian networks.

regards
Satnam