Abstraction for Belief Revision: Using a Genetic Algorithm to Compute the Most Probable Explanation

A belief network can create a compelling model of an agent’s uncertain environment. Exact belief network inference, including computing the most probable explanation, can be computationally hard. Therefore, it is interesting to perform inference on an approximate belief network rather than on the original belief network. This paper focuses on approximation in the form of abstraction. In particular, we show how a genetic algorithm can search for the most probable explanation in an abstracted belief network. Belief network approximation can be treated as noise from the point of view of a genetic algorithm, and there is therefore a relationship to research on noisy fitness functions used for genetic algorithms.

[1]  Sampath Srinivas,et al.  A Probabilistic Approach to Hierarchical Model-based Diagnosis , 1994, UAI.

[2]  Kazuhiro Saitou,et al.  Genetic algorithms as an approach to configuration and topology design , 1994, DAC 1993.

[3]  John J. Grefenstette,et al.  Genetic Search with Approximate Function Evaluation , 1985, ICGA.

[4]  Brad L. Miller,et al.  Noise, sampling, and efficient genetic algorthms , 1997 .

[5]  Kuo-Chu Chang,et al.  Refinement and coarsening of Bayesian networks , 1990, UAI.

[6]  Michael R. Genesereth,et al.  The Use of Design Descriptions in Automated Diagnosis , 1984, Artif. Intell..

[7]  Michael P. Wellman,et al.  On state-space abstraction for anytime evaluation of Bayesian networks , 1996, SGAR.

[8]  D. E. Goldberg,et al.  Genetic Algorithms in Search , 1989 .

[9]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems - networks of plausible inference , 1991, Morgan Kaufmann series in representation and reasoning.

[10]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[11]  Michael P. Wellman,et al.  State-Space Abstraction for Anytime Evaluation of Probabilistic Networks , 1994, UAI.

[12]  Victoria H. Clayton,et al.  Belief Network Algorithms : a Study of Performance Using Domain CharacterisationN , 1996 .

[13]  Ross D. Shachter.  Probabilistic Inference and Influence Diagrams , 1988, Oper. Res..

[14]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .