Paper Info
Title
Compiling bayesian networks for parameter learning based on shared BDDs
Abstract
Compiling Bayesian networks (BNs) is one of the most effective ways to exact inference because a logical approach enables the exploitation of local structures in BNs (i.e., determinism and context-specific independence). In this paper, a new parameter learning method based on compiling BNs is proposed. Firstly, a target BN with multiple evidence sets are compiled into a single shared binary decision diagram (SBDD) which shares common sub-graphs in multiple BDDs. Secondly, all conditional expectations which are required for executing the EM algorithm are simultaneously computed on the SBDD while their common local probabilities and expectations are shared. Due to these two types of sharing, the computation efficiency of the proposed method is higher than that of an EMalgorithm which naively uses an existing BN compiler for exact inference.
Year
DOI
Venue
2011
10.1007/978-3-642-25832-9_21
Australasian Conference on Artificial Intelligence
Keywords
Field
DocType
shared bdds,target bn,multiple evidence set,common sub-graphs,exact inference,multiple bdds,bayesian network,existing bn compiler,common local probability,single shared binary decision,local structure
Boolean function,Inference,Expectation–maximization algorithm,Computer science,Conditional expectation,Binary decision diagram,Compiler,Theoretical computer science,Bayesian network,Artificial intelligence,Machine learning,Computation
Conference
Volume
ISSN
Citations 
7106
0302-9743
0
PageRank 
References 
Authors
0.34
7
3
Name
Order
Citations
PageRank
Masakazu Ishihata1598.70
T. Sato21506137.10
Shin-ichi Minato372584.72