Efficient Algorithms for Bayesian Network Parameter Learning from Incomplete Data

Guy Van den Broeck Karthika Mohan Arthur Choi UCLA, Judea Pearl
Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence, PMLR R13:427-436, 2015.

Abstract

We propose a family of efficient algorithms for learning the parameters of a Bayesian network from incomplete data. Our approach is based on recent theoretical analyses of missing data problems, which utilize a graphical representation, called the missingness graph. In the case of MCAR and MAR data, this graph need not be explicit, and yet we can still obtain closed-form, asymptotically consistent parameter estimates, without the need for inference. When this missingness graph is explicated (based on background knowledge), even partially, we can obtain even more accurate estimates with less data. Empirically, we illustrate how we can learn the parameters of large networks from large datasets, which are beyond the scope of algorithms like EM (which require inference).

Cite this Paper

BibTeX
@InProceedings{pmlr-vR13-ucla15b, title = {Efficient Algorithms for {B}ayesian Network Parameter Learning from Incomplete Data}, author = {UCLA, Guy Van den Broeck Karthika Mohan Arthur Choi and Pearl, Judea}, booktitle = {Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence}, pages = {427--436}, year = {2015}, editor = {Meila, Marina and Heskes, Tom}, volume = {R13}, series = {Proceedings of Machine Learning Research}, month = {12--16 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r13/main/assets/ucla15b/ucla15b.pdf}, url = {https://proceedings.mlr.press/r13/ucla15b.html}, abstract = {We propose a family of efficient algorithms for learning the parameters of a Bayesian network from incomplete data. Our approach is based on recent theoretical analyses of missing data problems, which utilize a graphical representation, called the missingness graph. In the case of MCAR and MAR data, this graph need not be explicit, and yet we can still obtain closed-form, asymptotically consistent parameter estimates, without the need for inference. When this missingness graph is explicated (based on background knowledge), even partially, we can obtain even more accurate estimates with less data. Empirically, we illustrate how we can learn the parameters of large networks from large datasets, which are beyond the scope of algorithms like EM (which require inference).}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Efficient Algorithms for Bayesian Network Parameter Learning from Incomplete Data %A Guy Van den Broeck Karthika Mohan Arthur Choi UCLA %A Judea Pearl %B Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2015 %E Marina Meila %E Tom Heskes %F pmlr-vR13-ucla15b %I PMLR %P 427--436 %U https://proceedings.mlr.press/r13/ucla15b.html %V R13 %X We propose a family of efficient algorithms for learning the parameters of a Bayesian network from incomplete data. Our approach is based on recent theoretical analyses of missing data problems, which utilize a graphical representation, called the missingness graph. In the case of MCAR and MAR data, this graph need not be explicit, and yet we can still obtain closed-form, asymptotically consistent parameter estimates, without the need for inference. When this missingness graph is explicated (based on background knowledge), even partially, we can obtain even more accurate estimates with less data. Empirically, we illustrate how we can learn the parameters of large networks from large datasets, which are beyond the scope of algorithms like EM (which require inference). %Z Reissued by PMLR on 04 October 2026.
APA
UCLA, G.V.d.B.K.M.A.C. & Pearl, J.. (2015). Efficient Algorithms for Bayesian Network Parameter Learning from Incomplete Data. Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R13:427-436 Available from https://proceedings.mlr.press/r13/ucla15b.html. Reissued by PMLR on 04 October 2026.

Related Material