×

zbMATH — the first resource for mathematics

Toward better scoring metrics for pseudo-independent models. (English) Zbl 1101.68794
Summary: Learning belief networks from data is NP-hard in general. A common method used in heuristic learning is the single-link lookahead search. When the problem domain is Pseudo-Independent (PI), the method cannot discover the underlying probabilistic model. In learning these models, to explicitly trade model accuracy and model complexity, parameterization of PI models is necessary. Understanding of PI models also provides a new dimension of trade-off in learning even when the underlying model may not be PI. In this work, we adopt a hypercube perspective to analyze PI models and derive an improved result for computing the maximum number of parameters needed to specify a full PI model. We also present results on parameterization of a subclass of partial PI models.

MSC:
68T05 Learning and adaptive systems in artificial intelligence
68T35 Theory of languages and software systems (knowledge-based systems, expert systems, etc.) for artificial intelligence
Software:
WEBWEAVR-III
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] Lam, Comp Intell 10 pp 269– (1994)
[2] Cooper, Mach Learn 9 pp 309– (1992)
[3] Heckerman, Mach Learn 20 pp 197– (1995)
[4] Learning the structure of dynamic probabilistic networks. In: editors. Proc 14th Conf on Uncertainty in Artificial Intelligence. San Francisco, CA: Morgan Kaufmann; 1998. pp 139-147.
[5] Learning Bayesian networks: Search methods and experimental results. In: Proc of 5th Conf on Artificial Intelligence and Statistics. Society for AI and Statistics; 1995. pp 112-128.
[6] Critical remarks on single link search in learning belief networks. In: editors. Proc 12th Conf on Uncertainty in Artificial Intelligence. San Francisco, CA: Morgan Kaufmann; 1996. pp 564-571.
[7] Learning pseudo-independent models: Analytical and experimental results. In: editor. Advances in Artificial Intelligence. Berlin: Springer; 2000. pp 227-239. · Zbl 0987.68529
[8] Xiang, Mach Learn 26 pp 65– (1997)
[9] Learning belief networks in domains with recursively embedded pseudo independent submodels. In: editors. Proc 13th Conf on Uncertainty in Artificial Intelligence. San Francisco, CA: Morgan Kaufmann; 1997; pp 258-265.
[10] Towards understanding of pseudo-independent domains. In: editors. Poster Proc 10th Int Symp on Methodologies for Intelligent Systems. University of North Carolina-Charlotte; 1997.
[11] Probabilistic reasoning in intelligent systems: Networks of plausible inference. San Francisco, CA: Morgan Kaufmann, 1988.
[12] Graphical models. Oxford: Clarendon Press; 1996. · Zbl 0907.62001
[13] Probabilistic reasoning in multi-agent systems: A graphical models approach. Cambridge, UK: Cambridge University Press; 2002. · Zbl 1119.68198
[14] Kutato: An entropy-driven system for construction of probabilistic expert systems from database. In: editor. Proc 6th Conf on Uncertainty in Artificial Intelligence. Schenectady, NY: General Electric Corporate R&D; 1990. pp 54-62.
[15] Classifiers: A theoretical and empirical study. In: Proc 1991 Int Joint Conf on Artificial Intelligence; 1991. pp 638-644. · Zbl 0748.68046
[16] Madigan, J Amer Stat Assoc 89 pp 1535– (1994)
[17] Properties of Bayesian belief network learning algorithms. In: editors. Proc 10th Conf on Uncertainty in Artificial Intelligence. San Francisco, CA: Morgan Kaufmann; 1994. pp 102-109.
[18] Construction of a Markov network from data for probabilistic inference. In: editor. Proc 3rd Int Workshop on Rough Sets and Soft Computing. San Jose State University; 1994. pp 562-569.
[19] Overview of model selection. In: Proc 4th Int Workshop on Artificial Intelligence and Statistics. Society for AI and Statistics; 1993.
[20] Small-sample and large-sample statistical model selection criteria. In: editors. Selecting models from data. Berlin: Springer-Verlag; 1994. pp 31-39. · Zbl 0925.62028
[21] The recovery of causal ploy-trees from statistical data. In: editors. Proc of Workshop on Uncertainty in Artificial Intelligence. Amsterdam: Elsevier Science; 1987. pp 222-228.
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.