Carnegie Mellon University
Machine Learning Dept.
SELECT Lab

Home

 

 

 

 

 

 

We consider the problem of identifying a pattern of faults from a set of noisy linear measurements. Unfortunately, maximum a posteriori probability estimation of the fault pattern is computationally intractable. To solve the fault identification problem, we propose a non-parametric belief propagation approach. We show empirically that our belief propagation solver is more accurate than recent state-of-the-art algorithms including interior point methods and semidefinite programming. Our superior performance is explained by the fact that we take into account both the binary nature of the individual faults and the sparsity of the fault pattern arising from their rarity.

Papers

  • Fault identifiction via non-parametric belief propagation. D. Bickson, D. Baron, A. Ihler, H. Avissar and D. Dolev. IEEE Tran. on Signal Processing arxiv

    Talks

    Join us for Dror Baron's ITA 2011 talk. Sunday 2/06 - Friday 2/11, UCSD.

    Source Code

    Complete source code with all the experiments in the paper is found as part of the Gaussian belief propagation Matlab toolbox
    MMSE and BER bound computation code by Dror Baron.

  • Original site design by Pegasus Web Design Resources, modified by David Rabinowitz