Doubly Robust Data-Driven Distributionally Robust Optimization
Data-driven Distributionally Robust Optimization (DD-DRO) via optimal transport has been shown to encompass a wide range of popular machine learning algorithms. The distributional uncertainty size is often shown to correspond to the regularization parameter. The type of regularization (e.g. the norm...
Gespeichert in:
| Veröffentlicht in: | arXiv.org 2017-05 |
|---|---|
| Hauptverfasser: | , , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | arXiv.org |
| container_volume | |
| creator | Blanchet, Jose Kang, Yang Zhang, Fan He, Fei Hu, Zhangyi |
| description | Data-driven Distributionally Robust Optimization (DD-DRO) via optimal transport has been shown to encompass a wide range of popular machine learning algorithms. The distributional uncertainty size is often shown to correspond to the regularization parameter. The type of regularization (e.g. the norm used to regularize) corresponds to the shape of the distributional uncertainty. We propose a data-driven robust optimization methodology to inform the transportation cost underlying the definition of the distributional uncertainty. We show empirically that this additional layer of robustification, which produces a method we called doubly robust data-driven distributionally robust optimization (DD-R-DRO), allows to enhance the generalization properties of regularized estimators while reducing testing error relative to state-of-the-art classifiers in a wide range of data sets. |
| doi_str_mv | 10.48550/arxiv.1705.07168 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_arxiv</sourceid><recordid>TN_cdi_arxiv_primary_1705_07168</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2075602098</sourcerecordid><originalsourceid>FETCH-LOGICAL-a528-92f8a6e57822c8cac002877460a54cc8f9afe7bf9dac470f6e7ff08cd08abdce3</originalsourceid><addsrcrecordid>eNpFj11LwzAYhYMgOOZ-gFcWvG59mzbJ2xtBVr9gMJDdl7dpAhndWpN2OH-93SZ4deDwcDgPY3cpJDkKAY_kv90hSRWIBFQq8YrNeJalMeac37BFCFsA4FJxIbIZeyq7sW6P0WdXj2GIShooLr07mH1UujB4V4-D6_bU_jPrfnA790On_pZdW2qDWfzlnG1eXzbL93i1fvtYPq9iEhzjglskaYRCzjVq0tMBVCqXQCLXGm1B1qjaFg3pXIGVRlkLqBtAqhttsjm7v8ye5areux35Y3WSrM6SE_FwIXrffY0mDNW2G_30O1QclJDAocDsF-xzVbc</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2075602098</pqid></control><display><type>article</type><title>Doubly Robust Data-Driven Distributionally Robust Optimization</title><source>arXiv.org</source><source>Free E- Journals</source><creator>Blanchet, Jose ; Kang, Yang ; Zhang, Fan ; He, Fei ; Hu, Zhangyi</creator><creatorcontrib>Blanchet, Jose ; Kang, Yang ; Zhang, Fan ; He, Fei ; Hu, Zhangyi</creatorcontrib><description>Data-driven Distributionally Robust Optimization (DD-DRO) via optimal transport has been shown to encompass a wide range of popular machine learning algorithms. The distributional uncertainty size is often shown to correspond to the regularization parameter. The type of regularization (e.g. the norm used to regularize) corresponds to the shape of the distributional uncertainty. We propose a data-driven robust optimization methodology to inform the transportation cost underlying the definition of the distributional uncertainty. We show empirically that this additional layer of robustification, which produces a method we called doubly robust data-driven distributionally robust optimization (DD-R-DRO), allows to enhance the generalization properties of regularized estimators while reducing testing error relative to state-of-the-art classifiers in a wide range of data sets.</description><identifier>EISSN: 2331-8422</identifier><identifier>DOI: 10.48550/arxiv.1705.07168</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Algorithms ; Error reduction ; Machine learning ; Nonlinear programming ; Optimization ; Parameter uncertainty ; Regularization ; Robustness ; Statistics - Machine Learning</subject><ispartof>arXiv.org, 2017-05</ispartof><rights>2017. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,780,881,27902</link.rule.ids><backlink>$$Uhttps://doi.org/10.48550/arXiv.1705.07168$$DView paper in arXiv$$Hfree_for_read</backlink><backlink>$$Uhttps://doi.org/10.1002/9781119821588.ch4$$DView published paper (Access to full text may be restricted)$$Hfree_for_read</backlink></links><search><creatorcontrib>Blanchet, Jose</creatorcontrib><creatorcontrib>Kang, Yang</creatorcontrib><creatorcontrib>Zhang, Fan</creatorcontrib><creatorcontrib>He, Fei</creatorcontrib><creatorcontrib>Hu, Zhangyi</creatorcontrib><title>Doubly Robust Data-Driven Distributionally Robust Optimization</title><title>arXiv.org</title><description>Data-driven Distributionally Robust Optimization (DD-DRO) via optimal transport has been shown to encompass a wide range of popular machine learning algorithms. The distributional uncertainty size is often shown to correspond to the regularization parameter. The type of regularization (e.g. the norm used to regularize) corresponds to the shape of the distributional uncertainty. We propose a data-driven robust optimization methodology to inform the transportation cost underlying the definition of the distributional uncertainty. We show empirically that this additional layer of robustification, which produces a method we called doubly robust data-driven distributionally robust optimization (DD-R-DRO), allows to enhance the generalization properties of regularized estimators while reducing testing error relative to state-of-the-art classifiers in a wide range of data sets.</description><subject>Algorithms</subject><subject>Error reduction</subject><subject>Machine learning</subject><subject>Nonlinear programming</subject><subject>Optimization</subject><subject>Parameter uncertainty</subject><subject>Regularization</subject><subject>Robustness</subject><subject>Statistics - Machine Learning</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2017</creationdate><recordtype>article</recordtype><sourceid>BENPR</sourceid><sourceid>GOX</sourceid><recordid>eNpFj11LwzAYhYMgOOZ-gFcWvG59mzbJ2xtBVr9gMJDdl7dpAhndWpN2OH-93SZ4deDwcDgPY3cpJDkKAY_kv90hSRWIBFQq8YrNeJalMeac37BFCFsA4FJxIbIZeyq7sW6P0WdXj2GIShooLr07mH1UujB4V4-D6_bU_jPrfnA790On_pZdW2qDWfzlnG1eXzbL93i1fvtYPq9iEhzjglskaYRCzjVq0tMBVCqXQCLXGm1B1qjaFg3pXIGVRlkLqBtAqhttsjm7v8ye5areux35Y3WSrM6SE_FwIXrffY0mDNW2G_30O1QclJDAocDsF-xzVbc</recordid><startdate>20170519</startdate><enddate>20170519</enddate><creator>Blanchet, Jose</creator><creator>Kang, Yang</creator><creator>Zhang, Fan</creator><creator>He, Fei</creator><creator>Hu, Zhangyi</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope><scope>EPD</scope><scope>GOX</scope></search><sort><creationdate>20170519</creationdate><title>Doubly Robust Data-Driven Distributionally Robust Optimization</title><author>Blanchet, Jose ; Kang, Yang ; Zhang, Fan ; He, Fei ; Hu, Zhangyi</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a528-92f8a6e57822c8cac002877460a54cc8f9afe7bf9dac470f6e7ff08cd08abdce3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2017</creationdate><topic>Algorithms</topic><topic>Error reduction</topic><topic>Machine learning</topic><topic>Nonlinear programming</topic><topic>Optimization</topic><topic>Parameter uncertainty</topic><topic>Regularization</topic><topic>Robustness</topic><topic>Statistics - Machine Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Blanchet, Jose</creatorcontrib><creatorcontrib>Kang, Yang</creatorcontrib><creatorcontrib>Zhang, Fan</creatorcontrib><creatorcontrib>He, Fei</creatorcontrib><creatorcontrib>Hu, Zhangyi</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science & Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection><collection>arXiv Statistics</collection><collection>arXiv.org</collection><jtitle>arXiv.org</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Blanchet, Jose</au><au>Kang, Yang</au><au>Zhang, Fan</au><au>He, Fei</au><au>Hu, Zhangyi</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Doubly Robust Data-Driven Distributionally Robust Optimization</atitle><jtitle>arXiv.org</jtitle><date>2017-05-19</date><risdate>2017</risdate><eissn>2331-8422</eissn><abstract>Data-driven Distributionally Robust Optimization (DD-DRO) via optimal transport has been shown to encompass a wide range of popular machine learning algorithms. The distributional uncertainty size is often shown to correspond to the regularization parameter. The type of regularization (e.g. the norm used to regularize) corresponds to the shape of the distributional uncertainty. We propose a data-driven robust optimization methodology to inform the transportation cost underlying the definition of the distributional uncertainty. We show empirically that this additional layer of robustification, which produces a method we called doubly robust data-driven distributionally robust optimization (DD-R-DRO), allows to enhance the generalization properties of regularized estimators while reducing testing error relative to state-of-the-art classifiers in a wide range of data sets.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><doi>10.48550/arxiv.1705.07168</doi><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | EISSN: 2331-8422 |
| ispartof | arXiv.org, 2017-05 |
| issn | 2331-8422 |
| language | eng |
| recordid | cdi_arxiv_primary_1705_07168 |
| source | arXiv.org; Free E- Journals |
| subjects | Algorithms Error reduction Machine learning Nonlinear programming Optimization Parameter uncertainty Regularization Robustness Statistics - Machine Learning |
| title | Doubly Robust Data-Driven Distributionally Robust Optimization |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-30T00%3A50%3A04IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_arxiv&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Doubly%20Robust%20Data-Driven%20Distributionally%20Robust%20Optimization&rft.jtitle=arXiv.org&rft.au=Blanchet,%20Jose&rft.date=2017-05-19&rft.eissn=2331-8422&rft_id=info:doi/10.48550/arxiv.1705.07168&rft_dat=%3Cproquest_arxiv%3E2075602098%3C/proquest_arxiv%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2075602098&rft_id=info:pmid/&rfr_iscdi=true |