Abstract
Model complexity is an important factor to consider when selecting among graphical models. When all variables are observed, the complexity of a model can be measured by its standard dimension, i.e. the number of independent parameters. When latent variables are present, however, the standard dimension might no longer be appropriate. Instead, an effective dimension should be used [5]. Zhang & Kočka [13] showed how to compute the effective dimensions of partially observed trees. In this paper we solve the same problem for partially observed polytrees.
This is a preview of subscription content, log in via an institution to check access.
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
Akaike, H.: A new look at the statistical model identification. IEEE Transactions on Automatic Control AC-19, 716–723 (1974)
Catchpole, E.A., Morgan, B.J.T.: Deficiency of parameter-redundant models. Biometrika 88(2), 593–598 (2001)
Cheeseman, P., Stutz, J.: Bayesian classification (AutoClass): Theory and results. In: Fayyad, U., Paitesky-Shapiro, G., Smyth, P., Uthurusamy, R. (eds.) Advances in knowledge discovery and data mining, pp. 153–180 (1995)
Cooper, G., Herskovits, E.: A Bayesian method for the induction of probabilistic networks from data. Machine Learning 9, 309–347 (1992)
Geiger, D., Heckerman, D., Meek, C.: Asymptotic model selection for directed networks with hidden variables. In: Proc. of the 12th Conference on Uncertainty in Artificial Intelligence, pp. 283–290 (1996)
Kočka, T., Zhang, N.L.: Dimension correction for hierarchical latent class models. In: Proc. of the 18th Conference on Uncertainty in Artificial Intelligence (UAI 2002), pp. 267–274 (2002)
Lauritzen, S.L.: Graphical models. Clarendon Press, Oxford (1996)
Rusakov, D.: Effective dimension calculations for Bayesian networks, code in Matlab (2002), http://www.cs.technion.ac.il/~rusakov/archive/bn_dimension/bn_dimension.m
Rusakov, D., Geiger, D.: Asymptotic model Selection for naive Bayesian networks. In: Proc. of the 18th Conference on Uncertainty in Artificial Intelligence (UAI 2002), pp. 438–445 (2002)
Schwarz, G.: Estimating the dimension of a model. Annals of Statistics 6, 461–464 (1978)
Settimi, R., Smith, J.Q.: On the geometry of Bayesian graphical models with hidden variables. In: Proc. of the 14th Conference on Uncertainty in Artificial Intelligence, pp. 472–479 (1998)
Settimi, R., Smith, J.Q.: Geometry, moments and Bayesian networks with hidden variables. In: Proc. of the 7th International Workshop on Artificial Intelligence and Statistics, Fort Lauderdale, Florida, January 3–6. Morgan Kaufmann Publishers, San Francisco (1999)
Zhang, N.L., Kočka, T.: Effective dimensions of hierarchical latent class models, Technical Report HKUST-CS03-03, Department of Computer Science, Hong Kong University of Science and Technology (2003)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kočka, T., Zhang, N.L. (2003). Effective Dimensions of Partially Observed Polytrees. In: Nielsen, T.D., Zhang, N.L. (eds) Symbolic and Quantitative Approaches to Reasoning with Uncertainty. ECSQARU 2003. Lecture Notes in Computer Science(), vol 2711. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-45062-7_15
Download citation
DOI: https://doi.org/10.1007/978-3-540-45062-7_15
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-40494-1
Online ISBN: 978-3-540-45062-7
eBook Packages: Springer Book Archive