Title :
Approximating Bayesian belief networks by arc removal
Author :
Van Engelen, Robert A.
Author_Institution :
Dept. of Comput. Sci., Leiden Univ., Netherlands
fDate :
8/1/1997 12:00:00 AM
Abstract :
I propose a general framework for approximating Bayesian belief networks through model simplification by arc removal. Given an upper bound on the absolute error allowed on the prior and posterior probability distributions of the approximated network, a subset of arcs is removed, thereby speeding up probabilistic inference
Keywords :
directed graphs; inference mechanisms; probability; uncertainty handling; Bayesian belief networks; absolute error; arc removal; model simplification; posterior probability distribution; prior probability distribution; probabilistic inference; Application software; Bayesian methods; Computational modeling; Decision making; Frequency estimation; Information theory; Medical diagnosis; Probability distribution; Uncertainty; Upper bound;
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
