Abstract
This paper addresses an important issue about the compact representation of the conditional probability distribution (CPD) applied in the well known Bayesian Networks in uncertain causality representation and probabilistic inference. That is, there is an essential difference between the single-valued cases and the multi-valued cases, while this difference does not exist when the CPD is represented in the conditional probability table (CPT). In other words, the present compact representation and inference methods applicable in the single-valued cases may not be applicable in the multi-valued cases as people usually think. A detailed example is provided to illustrate this problem. The solution is provided in the references by the author.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Pearl, J.: Probabilistic reasoning in intelligent systems. Morgan Kaufmann, San Mateo (1988)
Boutilier, C., Friedman, N., Goldszmidt, M., Koller, D.: Context-specific independence in Bayesian Network. In: Proc. of UAI 1996 (1996)
Zhang, Q.: Probabilistic reasoning based on dynamic causality trees/diagrams. Reliability Engineering and System Safety (46), 209–220 (1994)
Poole, D., Zhang, N.L.: Exploiting contextual independence in probabilistic inference. Journal of Artificial Intelligence Research 18, 263–313 (2003)
D’Ambrosio, B.: Local expression languages for probabilistic dependence. Int. J. Approximate Reasoning 139(1), 61–68 (1995)
Zhang, Q.: DUCG - a new approach to deal with dynamical uncertain causality information - part I: the static discrete DAG case. Submitted for publication in IEEE Trans. System, Man and Cybernetics - Part A: Systems and Human
Zhang, Q.: DUCG - a new approach to deal with dynamical uncertain causality information - part II: the dynamical DCG case. Submitted for publication in IEEE Trans. System, Man and Cybernetics - Part A: Systems and Human
Zhang, Q.: DUCG - a new approach to deal with dynamical uncertain causality information - part III: the discrete dynamical case. Submitted for publication in IEEE Trans. System, Man and Cybernetics - Part A: Systems and Human
Zhang, Q.: DUCG - a new approach to deal with dynamical uncertain causality information - part IV: The Continuous Variable and Uncertain Evidence. Submitted for publication in IEEE Trans. System, Man and Cybernetics - Part A: Systems and Human
Zhang, Q.: An application of DUCG – The dynamical fault diagnoses and predictions of a nuclear power plant. Submitted for publication in IEEE Trans. System, Man and Cybernetics - Part A: Systems and Human
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 ICST Institute for Computer Science, Social Informatics and Telecommunications Engineering
About this paper
Cite this paper
Zhang, Q. (2009). The Difference between Single-Valued and Multi-Valued Cases in the Compact Representation of CPD in Bayesian Networks. In: Zhou, J. (eds) Complex Sciences. Complex 2009. Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, vol 4. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-02466-5_58
Download citation
DOI: https://doi.org/10.1007/978-3-642-02466-5_58
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-02465-8
Online ISBN: 978-3-642-02466-5
eBook Packages: Computer ScienceComputer Science (R0)