Entropic optimal transport is maximum-likelihood deconvolution
We give a statistical interpretation of entropic optimal transport by showing that performing maximum-likelihood estimation for Gaussian deconvolution corresponds to calculating a projection with respect to the entropic optimal transport distance. This structural result gives theoretical support for...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Elsevier BV
2020
|
| Online Access: | https://hdl.handle.net/1721.1/126692 |
| _version_ | 1811083480927305728 |
|---|---|
| author | Rigollet, Philippe Weed, Jonathan |
| author2 | Massachusetts Institute of Technology. Department of Mathematics |
| author_facet | Massachusetts Institute of Technology. Department of Mathematics Rigollet, Philippe Weed, Jonathan |
| author_sort | Rigollet, Philippe |
| collection | MIT |
| description | We give a statistical interpretation of entropic optimal transport by showing that performing maximum-likelihood estimation for Gaussian deconvolution corresponds to calculating a projection with respect to the entropic optimal transport distance. This structural result gives theoretical support for the wide adoption of these tools in the machine learning community. |
| first_indexed | 2024-09-23T12:33:45Z |
| format | Article |
| id | mit-1721.1/126692 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T12:33:45Z |
| publishDate | 2020 |
| publisher | Elsevier BV |
| record_format | dspace |
| spelling | mit-1721.1/1266922022-09-28T08:39:10Z Entropic optimal transport is maximum-likelihood deconvolution Rigollet, Philippe Weed, Jonathan Massachusetts Institute of Technology. Department of Mathematics We give a statistical interpretation of entropic optimal transport by showing that performing maximum-likelihood estimation for Gaussian deconvolution corresponds to calculating a projection with respect to the entropic optimal transport distance. This structural result gives theoretical support for the wide adoption of these tools in the machine learning community. 2020-08-20T00:59:20Z 2020-08-20T00:59:20Z 2018-11 2019-11-19T17:33:20Z Article http://purl.org/eprint/type/JournalArticle 1631-073X https://hdl.handle.net/1721.1/126692 Rigollet, Philippe and Jonathan Weed. "Entropic optimal transport is maximum-likelihood deconvolution." Comptes Rendus Mathematique 356, 11-12 (November 2018): 1228-1235 © 2018 Académie des sciences en http://dx.doi.org/10.1016/j.crma.2018.10.010 Comptes Rendus Mathematique Creative Commons Attribution-NonCommercial-NoDerivs License http://creativecommons.org/licenses/by-nc-nd/4.0/ application/pdf Elsevier BV arXiv |
| spellingShingle | Rigollet, Philippe Weed, Jonathan Entropic optimal transport is maximum-likelihood deconvolution |
| title | Entropic optimal transport is maximum-likelihood deconvolution |
| title_full | Entropic optimal transport is maximum-likelihood deconvolution |
| title_fullStr | Entropic optimal transport is maximum-likelihood deconvolution |
| title_full_unstemmed | Entropic optimal transport is maximum-likelihood deconvolution |
| title_short | Entropic optimal transport is maximum-likelihood deconvolution |
| title_sort | entropic optimal transport is maximum likelihood deconvolution |
| url | https://hdl.handle.net/1721.1/126692 |
| work_keys_str_mv | AT rigolletphilippe entropicoptimaltransportismaximumlikelihooddeconvolution AT weedjonathan entropicoptimaltransportismaximumlikelihooddeconvolution |