Distributed Parallel Inference on Large Factor Graphs

Joseph Gonzalez, Yucheng Low, Carlos Guestrin, David O’Hallaron
Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence, PMLR R7:203-212, 2009.

Abstract

As computer clusters become more common and the size of the problems encountered in the field of AI grows, there is an increasing demand for efficient parallel inference algorithms. We consider the problem of parallel inference on large factor graphs in the distributed memory setting of computer clusters. We develop a new efficient parallel inference algorithm, DBRSplash, which incorporates over-segmented graph partitioning, belief residual scheduling, and uniform work Splash operations. We empirically evaluate the DBRSplash algorithm on a 120 processor cluster and demonstrate linear to super-linear performance gains on large factor graph models.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR7-gonzalez09a, title = {Distributed Parallel Inference on Large Factor Graphs}, author = {Gonzalez, Joseph and Low, Yucheng and Guestrin, Carlos and O'Hallaron, David}, booktitle = {Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence}, pages = {203--212}, year = {2009}, editor = {Bilmes, Jeff and Ng, Andrew Y.}, volume = {R7}, series = {Proceedings of Machine Learning Research}, month = {18--21 Jun}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r7/main/assets/gonzalez09a/gonzalez09a.pdf}, url = {https://proceedings.mlr.press/r7/gonzalez09a.html}, abstract = {As computer clusters become more common and the size of the problems encountered in the field of AI grows, there is an increasing demand for efficient parallel inference algorithms. We consider the problem of parallel inference on large factor graphs in the distributed memory setting of computer clusters. We develop a new efficient parallel inference algorithm, DBRSplash, which incorporates over-segmented graph partitioning, belief residual scheduling, and uniform work Splash operations. We empirically evaluate the DBRSplash algorithm on a 120 processor cluster and demonstrate linear to super-linear performance gains on large factor graph models.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Distributed Parallel Inference on Large Factor Graphs %A Joseph Gonzalez %A Yucheng Low %A Carlos Guestrin %A David O’Hallaron %B Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2009 %E Jeff Bilmes %E Andrew Y. Ng %F pmlr-vR7-gonzalez09a %I PMLR %P 203--212 %U https://proceedings.mlr.press/r7/gonzalez09a.html %V R7 %X As computer clusters become more common and the size of the problems encountered in the field of AI grows, there is an increasing demand for efficient parallel inference algorithms. We consider the problem of parallel inference on large factor graphs in the distributed memory setting of computer clusters. We develop a new efficient parallel inference algorithm, DBRSplash, which incorporates over-segmented graph partitioning, belief residual scheduling, and uniform work Splash operations. We empirically evaluate the DBRSplash algorithm on a 120 processor cluster and demonstrate linear to super-linear performance gains on large factor graph models. %Z Reissued by PMLR on 04 October 2026.
APA
Gonzalez, J., Low, Y., Guestrin, C. & O’Hallaron, D.. (2009). Distributed Parallel Inference on Large Factor Graphs. Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R7:203-212 Available from https://proceedings.mlr.press/r7/gonzalez09a.html. Reissued by PMLR on 04 October 2026.

Related Material