Scalable MAP inference in Bayesian networks based on a Map-Reduce approach

Darı́o Ramos-López, Antonio Salmerón, Rafel Rumı́, Ana M. Martı́nez, Thomas D. Nielsen, Andrés R. Masegosa, Helge Langseth, Anders L. Madsen
; Proceedings of the Eighth International Conference on Probabilistic Graphical Models, PMLR 52:415-425, 2016.

Abstract

Maximum a posteriori (MAP) inference is a particularly complex type of probabilistic inference in Bayesian networks. It consists of finding the most probable configuration of a set of variables of interest given observations on a collection of other variables. In this paper we study scalable solutions to the MAP problem in hybrid Bayesian networks parameterized using conditional linear Gaussian distributions. We propose scalable solutions based on hill climbing and simulated annealing, built on the Apache Flink framework for big data processing. We analyze the scalability of the solution through a series of experiments on large synthetic networks.

Cite this Paper


BibTeX
@InProceedings{pmlr-v52-ramos-lopez16, title = {Scalable MAP inference in {B}ayesian networks based on a Map-Reduce approach}, author = {Darı́o Ramos-López and Antonio Salmerón and Rafel Rumı́ and Ana M. Martı́nez and Thomas D. Nielsen and Andrés R. Masegosa and Helge Langseth and Anders L. Madsen}, booktitle = {Proceedings of the Eighth International Conference on Probabilistic Graphical Models}, pages = {415--425}, year = {2016}, editor = {Alessandro Antonucci and Giorgio Corani and Cassio Polpo Campos}}, volume = {52}, series = {Proceedings of Machine Learning Research}, address = {Lugano, Switzerland}, month = {06--09 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v52/ramos-lopez16.pdf}, url = {http://proceedings.mlr.press/v52/ramos-lopez16.html}, abstract = {Maximum a posteriori (MAP) inference is a particularly complex type of probabilistic inference in Bayesian networks. It consists of finding the most probable configuration of a set of variables of interest given observations on a collection of other variables. In this paper we study scalable solutions to the MAP problem in hybrid Bayesian networks parameterized using conditional linear Gaussian distributions. We propose scalable solutions based on hill climbing and simulated annealing, built on the Apache Flink framework for big data processing. We analyze the scalability of the solution through a series of experiments on large synthetic networks.} }
Endnote
%0 Conference Paper %T Scalable MAP inference in Bayesian networks based on a Map-Reduce approach %A Darı́o Ramos-López %A Antonio Salmerón %A Rafel Rumı́ %A Ana M. Martı́nez %A Thomas D. Nielsen %A Andrés R. Masegosa %A Helge Langseth %A Anders L. Madsen %B Proceedings of the Eighth International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2016 %E Alessandro Antonucci %E Giorgio Corani %E Cassio Polpo Campos} %F pmlr-v52-ramos-lopez16 %I PMLR %J Proceedings of Machine Learning Research %P 415--425 %U http://proceedings.mlr.press %V 52 %W PMLR %X Maximum a posteriori (MAP) inference is a particularly complex type of probabilistic inference in Bayesian networks. It consists of finding the most probable configuration of a set of variables of interest given observations on a collection of other variables. In this paper we study scalable solutions to the MAP problem in hybrid Bayesian networks parameterized using conditional linear Gaussian distributions. We propose scalable solutions based on hill climbing and simulated annealing, built on the Apache Flink framework for big data processing. We analyze the scalability of the solution through a series of experiments on large synthetic networks.
RIS
TY - CPAPER TI - Scalable MAP inference in Bayesian networks based on a Map-Reduce approach AU - Darı́o Ramos-López AU - Antonio Salmerón AU - Rafel Rumı́ AU - Ana M. Martı́nez AU - Thomas D. Nielsen AU - Andrés R. Masegosa AU - Helge Langseth AU - Anders L. Madsen BT - Proceedings of the Eighth International Conference on Probabilistic Graphical Models PY - 2016/08/15 DA - 2016/08/15 ED - Alessandro Antonucci ED - Giorgio Corani ED - Cassio Polpo Campos} ID - pmlr-v52-ramos-lopez16 PB - PMLR SP - 415 DP - PMLR EP - 425 L1 - http://proceedings.mlr.press/v52/ramos-lopez16.pdf UR - http://proceedings.mlr.press/v52/ramos-lopez16.html AB - Maximum a posteriori (MAP) inference is a particularly complex type of probabilistic inference in Bayesian networks. It consists of finding the most probable configuration of a set of variables of interest given observations on a collection of other variables. In this paper we study scalable solutions to the MAP problem in hybrid Bayesian networks parameterized using conditional linear Gaussian distributions. We propose scalable solutions based on hill climbing and simulated annealing, built on the Apache Flink framework for big data processing. We analyze the scalability of the solution through a series of experiments on large synthetic networks. ER -
APA
Ramos-López, D., Salmerón, A., Rumı́, R., Martı́nez, A.M., Nielsen, T.D., Masegosa, A.R., Langseth, H. & Madsen, A.L.. (2016). Scalable MAP inference in Bayesian networks based on a Map-Reduce approach. Proceedings of the Eighth International Conference on Probabilistic Graphical Models, in PMLR 52:415-425

Related Material