Approximate Decentralized Bayesian Inference
URL to accepted papers on conference site
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Association for Uncertainty in Artificial Intelligence Press
2014
|
| Online Access: | http://hdl.handle.net/1721.1/88106 https://orcid.org/0000-0003-1499-0191 https://orcid.org/0000-0001-8576-1930 |
| _version_ | 1811073643484020736 |
|---|---|
| author | Campbell, Trevor David How, Jonathan P. |
| author2 | Massachusetts Institute of Technology. Department of Aeronautics and Astronautics |
| author_facet | Massachusetts Institute of Technology. Department of Aeronautics and Astronautics Campbell, Trevor David How, Jonathan P. |
| author_sort | Campbell, Trevor David |
| collection | MIT |
| description | URL to accepted papers on conference site |
| first_indexed | 2024-09-23T09:36:17Z |
| format | Article |
| id | mit-1721.1/88106 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T09:36:17Z |
| publishDate | 2014 |
| publisher | Association for Uncertainty in Artificial Intelligence Press |
| record_format | dspace |
| spelling | mit-1721.1/881062022-09-30T15:36:05Z Approximate Decentralized Bayesian Inference Campbell, Trevor David How, Jonathan P. Massachusetts Institute of Technology. Department of Aeronautics and Astronautics Campbell, Trevor David Campbell, Trevor David How, Jonathan P. URL to accepted papers on conference site This paper presents an approximate method for performing Bayesian inference in models with conditional independence over a decentralized network of learning agents. The method first employs variational inference on each individual learning agent to generate a local approximate posterior, the agents transmit their local posteriors to other agents in the network, and finally each agent combines its set of received local posteriors. The key insight in this work is that, for many Bayesian models, approximate inference schemes destroy symmetry and dependencies in the model that are crucial to the correct application of Bayes’ rule when combining the local posteriors. The proposed method addresses this issue by including an additional optimization step in the combination procedure that accounts for these broken dependencies. Experiments on synthetic and real data demonstrate that the decentralized method provides advantages in computational performance and predictive test likelihood over previous batch and distributed methods. United States. Office of Naval Research (ONR MURI grant N000141110688) 2014-06-27T17:42:15Z 2014-06-27T17:42:15Z 2014-07 Article http://purl.org/eprint/type/ConferencePaper ID: 182 http://hdl.handle.net/1721.1/88106 Campbell, Trevor and Jonathan How. "Approximate Decentralized Bayesian Inference." 30th Conference on Uncertainty in Artificial Intelligence, UAI 2014, Quebec City, Quebec, Canada, July 23-27, 2014. p.1-10. https://orcid.org/0000-0003-1499-0191 https://orcid.org/0000-0001-8576-1930 en_US http://auai.org/uai2014/acceptedPapers.shtml Proceedings of the 30th Conference on Uncertainty in Artificial Intelligence, UAI 2014 Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Association for Uncertainty in Artificial Intelligence Press arXiv |
| spellingShingle | Campbell, Trevor David How, Jonathan P. Approximate Decentralized Bayesian Inference |
| title | Approximate Decentralized Bayesian Inference |
| title_full | Approximate Decentralized Bayesian Inference |
| title_fullStr | Approximate Decentralized Bayesian Inference |
| title_full_unstemmed | Approximate Decentralized Bayesian Inference |
| title_short | Approximate Decentralized Bayesian Inference |
| title_sort | approximate decentralized bayesian inference |
| url | http://hdl.handle.net/1721.1/88106 https://orcid.org/0000-0003-1499-0191 https://orcid.org/0000-0001-8576-1930 |
| work_keys_str_mv | AT campbelltrevordavid approximatedecentralizedbayesianinference AT howjonathanp approximatedecentralizedbayesianinference |