Results 1 
2 of
2
Learning Bayesian belief networks: An approach based on the MDL principle
 Computational Intelligence
, 1994
"... A new approach for learning Bayesian belief networks from raw data is presented. The approach is based on Rissanen's Minimal Description Length (MDL) principle, which is particularly well suited for this task. Our approach does not require any prior assumptions about the distribution being lear ..."
Abstract

Cited by 247 (7 self)
 Add to MetaCart
(Show Context)
A new approach for learning Bayesian belief networks from raw data is presented. The approach is based on Rissanen's Minimal Description Length (MDL) principle, which is particularly well suited for this task. Our approach does not require any prior assumptions about the distribution being learned. In particular, our method can learn unrestricted multiplyconnected belief networks. Furthermore, unlike other approaches our method allows us to tradeo accuracy and complexity in the learned model. This is important since if the learned model is very complex (highly connected) it can be conceptually and computationally intractable. In such a case it would be preferable to use a simpler model even if it is less accurate. The MDL principle o ers a reasoned method for making this tradeo. We also show that our method generalizes previous approaches based on Kullback crossentropy. Experiments have been conducted to demonstrate the feasibility of the approach. Keywords: Knowledge Acquisition � Bayes Nets � Uncertainty Reasoning. 1
Bounded RD
"... This paper presents a new inference algorithm for belief networks that combines a searchbased algorithm with a simulationbased algorithm. The former is an extension of the recursive decomposition (RD) algorithm proposed by Cooper in [8], which is here modified to compute interval bounds on margina ..."
Abstract
 Add to MetaCart
(Show Context)
This paper presents a new inference algorithm for belief networks that combines a searchbased algorithm with a simulationbased algorithm. The former is an extension of the recursive decomposition (RD) algorithm proposed by Cooper in [8], which is here modified to compute interval bounds on marginal probabilities. We call the algorithm boundedRD. The latter is a stochastic simulation method known as Pearl's Markov blanket algorithm [31]. Markov simulation is used to generate highly probable instantiations of the network nodes to be used by boundedRD in the computation of probability bounds. BoundedRD has the anytime property, and produces successively narrower interval bounds, which converge in the limit to the exact value.