Abstract
We consider probabilistic predictions using graphical models and describe a newly developed method, fully conditional Venn predictor (FCVP). FCVP can provide upper and lower bounds for the conditional probability associated with each predicted label. Empirical results confirm that FCVP can give well-calibrated predictions in online learning mode. Experimental results also show the prediction performance of FCVP is good in both the online and the offline learning setting without making any additional assumptions, apart from i.i.d.
Access provided by Autonomous University of Puebla. Download to read the full chapter text
Chapter PDF
Similar content being viewed by others
References
Cowell, R.G., Dawid, A.P., Lauritzen, S.L., Spiegelhalter, D.J.: Probabilistic Networks and Expert Systems. Statistics for Engineering and Information Science. Springer, Heidelberg (1999)
Heckerman, D., Geiger, D.: Likelihoods and parameter priors for bayesian networks. Technical Report MSR-TR-95-54, Microsoft Research (1995)
Finn, V.: Jensen: An introduction to Bayesian Networks. Taylor and Francis, London (1996)
Lauritzen, S.L., Spiegelhalter, D.J.: Local computations with probabilities on graphical structures and their application to expert systems (with discussion). J. Royal Statist. Soc. B(50), 157–224 (1988)
Luo, Z., Gammerman, A.: Parameter learning in Bayesian belief networks. In: Proceeding of IPMU 1992, pp. 25–28 (1992)
Murphy, K.: The Bayes Net Toolbox for Matlab. Computing Science and Statistics 33 (2001)
Spiegelhalter, D.J., Lauritzen, S.L.: Sequential updating of conditional probabilities on directed graphical structures. Networks 20(5), 579–605 (1990)
Vovk, V., Shafer, G., Nouretdinov, I.: Self-calibrating probability forecasting. In: Thrun, S., Saul, L., Schölkopf, B. (eds.) Advances in Neural Information Processing Systems 16, MIT Press, Cambridge (2004)
Vovk, V., Gammerman, A., Shafer, G.: Algorithmic learning in a random world. Springer, Heidelberg (2005) (to appear)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Luo, Z., Gammerman, A. (2005). Qualified Probabilistic Predictions Using Graphical Models. In: Godo, L. (eds) Symbolic and Quantitative Approaches to Reasoning with Uncertainty. ECSQARU 2005. Lecture Notes in Computer Science(), vol 3571. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11518655_11
Download citation
DOI: https://doi.org/10.1007/11518655_11
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-27326-4
Online ISBN: 978-3-540-31888-0
eBook Packages: Computer ScienceComputer Science (R0)