Exploiting Within-Clique Factorizations in Max-Product Message-Passing Algorithms
Julian McAuley (SoCS CECS)
CS HDR MONITORING AI Research GroupDATE: 2010-04-15
TIME: 09:00:00 - 09:30:00
LOCATION: RSISE Seminar Room, ground floor, building 115, cnr. North and Daley Roads, ANU
CONTACT: JavaScript must be enabled to display this email address.
ABSTRACT:
In our previous talks, we presented algorithms that improve the performance of exact and approximate max-product message-passing for models with factors smaller than their maximal cliques, and more generally, for models whose factors that depend upon the observation contain fewer latent variables than their maximal cliques. In this talk, we present a variety of applications to which these algorithms can be applied, and suggest a variety of extensions that build upon our previous results.
BIO:
PhD student in Computer Science.
