PMID- 20740057
OWN - NLM
STAT- PubMed-not-MEDLINE
LR  - 20240318
IS  - 1076-9757 (Print)
IS  - 1943-5037 (Electronic)
IS  - 1076-9757 (Linking)
VI  - 37
DP  - 2010 Jan 1
TI  - Join-Graph Propagation Algorithms.
PG  - 279-328
AB  - The paper investigates parameterized approximate message-passing schemes that are 
      based on bounded inference and are inspired by Pearl's belief propagation 
      algorithm (BP). We start with the bounded inference mini-clustering algorithm and 
      then move to the iterative scheme called Iterative Join-Graph Propagation (IJGP), 
      that combines both iteration and bounded inference. Algorithm IJGP belongs to the 
      class of Generalized Belief Propagation algorithms, a framework that allowed 
      connections with approximate algorithms from statistical physics and is shown 
      empirically to surpass the performance of mini-clustering and belief propagation, 
      as well as a number of other state-of-the-art algorithms on several classes of 
      networks. We also provide insight into the accuracy of iterative BP and IJGP by 
      relating these algorithms to well known classes of constraint propagation 
      schemes.
FAU - Mateescu, Robert
AU  - Mateescu R
AD  - Microsoft Research 7 J J Thomson Avenue Cambridge CB3 0FB, UK 
      ROMATEES@MICROSOFT.COM.
FAU - Kask, Kalev
AU  - Kask K
FAU - Gogate, Vibhav
AU  - Gogate V
FAU - Dechter, Rina
AU  - Dechter R
LA  - eng
GR  - R01 HG004175/HG/NHGRI NIH HHS/United States
GR  - R01 HG004175-02/HG/NHGRI NIH HHS/United States
PT  - Journal Article
PL  - United States
TA  - J Artif Intell Res
JT  - The journal of artificial intelligence research
JID - 101512254
PMC - PMC2926991
MID - NIHMS220551
EDAT- 2010/08/27 06:00
MHDA- 2010/08/27 06:01
PMCR- 2010/08/24
CRDT- 2010/08/27 06:00
PHST- 2010/08/27 06:00 [entrez]
PHST- 2010/08/27 06:00 [pubmed]
PHST- 2010/08/27 06:01 [medline]
PHST- 2010/08/24 00:00 [pmc-release]
AID - 10.1613/jair.2842 [doi]
PST - ppublish
SO  - J Artif Intell Res. 2010 Jan 1;37:279-328. doi: 10.1613/jair.2842.