Sublinear Optimization for Machine Learning
In this article we describe and analyze sublinear-time approximation algorithms for some optimization problems arising in machine learning, such as training linear classifiers and finding minimum enclosing balls. Our algorithms can be extended to some kernelized versions of these problems, such as S...
Gespeichert in:
| Veröffentlicht in: | Journal of the ACM 2012-10, Vol.59 (5), p.1-49 |
|---|---|
| Hauptverfasser: | , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 49 |
|---|---|
| container_issue | 5 |
| container_start_page | 1 |
| container_title | Journal of the ACM |
| container_volume | 59 |
| creator | CLARKSON, Kenneth L HAZAN, Elad WOODRUFF, David P |
| description | In this article we describe and analyze sublinear-time approximation algorithms for some optimization problems arising in machine learning, such as training linear classifiers and finding minimum enclosing balls. Our algorithms can be extended to some kernelized versions of these problems, such as SVDD, hard margin SVM, and
L
2
-SVM, for which sublinear-time algorithms were not known before. These new algorithms use a combination of a novel sampling techniques and a new multiplicative update algorithm. We give lower bounds which show the running times of many of our algorithms to be nearly best possible in the unit-cost RAM model. |
| doi_str_mv | 10.1145/2371656.2371658 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_1506361775</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2818926901</sourcerecordid><originalsourceid>FETCH-LOGICAL-c332t-6d11601e123c8defeb2e00a5e94863cba0d5ff3180bc8e2e52c1a5575b6ddd8e3</originalsourceid><addsrcrecordid>eNpdkM1LxDAQxYMouK6evRZEEKS7M0nz4VGW9QNW9qCCt5KmqWbppmvSHvSvN7LFgzDwGN7vDcMj5BxhhljwOWUSBRezvaoDMkHOZS4ZfzskEwAocl4gHpOTGDdpBQpyQq6fh6p13uqQrXe927pv3bvOZ00XsidtPpKVrZLrnX8_JUeNbqM9G3VKXu-WL4uHfLW-f1zcrnLDGO1zUSMKQIuUGVXbxlbUAmhubwolmKk01LxpGCqojLLUcmpQp1d5Jeq6VpZNydX-7i50n4ONfbl10di21d52QyyRg2ACpeQJvfiHbroh-PRdiSghQWkSNd9TJnQxBtuUu-C2OnyVCOVveeVY3qgqJS7Huzoa3TZBe-PiX4wKCSiRsR_4mmyD</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>1170617617</pqid></control><display><type>article</type><title>Sublinear Optimization for Machine Learning</title><source>ACM Digital Library Complete</source><creator>CLARKSON, Kenneth L ; HAZAN, Elad ; WOODRUFF, David P</creator><creatorcontrib>CLARKSON, Kenneth L ; HAZAN, Elad ; WOODRUFF, David P</creatorcontrib><description>In this article we describe and analyze sublinear-time approximation algorithms for some optimization problems arising in machine learning, such as training linear classifiers and finding minimum enclosing balls. Our algorithms can be extended to some kernelized versions of these problems, such as SVDD, hard margin SVM, and
L
2
-SVM, for which sublinear-time algorithms were not known before. These new algorithms use a combination of a novel sampling techniques and a new multiplicative update algorithm. We give lower bounds which show the running times of many of our algorithms to be nearly best possible in the unit-cost RAM model.</description><identifier>ISSN: 0004-5411</identifier><identifier>EISSN: 1557-735X</identifier><identifier>DOI: 10.1145/2371656.2371658</identifier><identifier>CODEN: JACOAH</identifier><language>eng</language><publisher>New York, NY: Association for Computing Machinery</publisher><subject>Algorithmics. Computability. Computer arithmetics ; Algorithms ; Applied sciences ; Artificial intelligence ; Branch & bound algorithms ; Computer science; control theory; systems ; Exact sciences and technology ; Linear programming ; Machine learning ; Optimization algorithms ; Studies ; Theoretical computing</subject><ispartof>Journal of the ACM, 2012-10, Vol.59 (5), p.1-49</ispartof><rights>2014 INIST-CNRS</rights><rights>Copyright Association for Computing Machinery Oct 2012</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c332t-6d11601e123c8defeb2e00a5e94863cba0d5ff3180bc8e2e52c1a5575b6ddd8e3</citedby><cites>FETCH-LOGICAL-c332t-6d11601e123c8defeb2e00a5e94863cba0d5ff3180bc8e2e52c1a5575b6ddd8e3</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>314,780,784,27924,27925</link.rule.ids><backlink>$$Uhttp://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=26701713$$DView record in Pascal Francis$$Hfree_for_read</backlink></links><search><creatorcontrib>CLARKSON, Kenneth L</creatorcontrib><creatorcontrib>HAZAN, Elad</creatorcontrib><creatorcontrib>WOODRUFF, David P</creatorcontrib><title>Sublinear Optimization for Machine Learning</title><title>Journal of the ACM</title><description>In this article we describe and analyze sublinear-time approximation algorithms for some optimization problems arising in machine learning, such as training linear classifiers and finding minimum enclosing balls. Our algorithms can be extended to some kernelized versions of these problems, such as SVDD, hard margin SVM, and
L
2
-SVM, for which sublinear-time algorithms were not known before. These new algorithms use a combination of a novel sampling techniques and a new multiplicative update algorithm. We give lower bounds which show the running times of many of our algorithms to be nearly best possible in the unit-cost RAM model.</description><subject>Algorithmics. Computability. Computer arithmetics</subject><subject>Algorithms</subject><subject>Applied sciences</subject><subject>Artificial intelligence</subject><subject>Branch & bound algorithms</subject><subject>Computer science; control theory; systems</subject><subject>Exact sciences and technology</subject><subject>Linear programming</subject><subject>Machine learning</subject><subject>Optimization algorithms</subject><subject>Studies</subject><subject>Theoretical computing</subject><issn>0004-5411</issn><issn>1557-735X</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2012</creationdate><recordtype>article</recordtype><recordid>eNpdkM1LxDAQxYMouK6evRZEEKS7M0nz4VGW9QNW9qCCt5KmqWbppmvSHvSvN7LFgzDwGN7vDcMj5BxhhljwOWUSBRezvaoDMkHOZS4ZfzskEwAocl4gHpOTGDdpBQpyQq6fh6p13uqQrXe927pv3bvOZ00XsidtPpKVrZLrnX8_JUeNbqM9G3VKXu-WL4uHfLW-f1zcrnLDGO1zUSMKQIuUGVXbxlbUAmhubwolmKk01LxpGCqojLLUcmpQp1d5Jeq6VpZNydX-7i50n4ONfbl10di21d52QyyRg2ACpeQJvfiHbroh-PRdiSghQWkSNd9TJnQxBtuUu-C2OnyVCOVveeVY3qgqJS7Huzoa3TZBe-PiX4wKCSiRsR_4mmyD</recordid><startdate>20121001</startdate><enddate>20121001</enddate><creator>CLARKSON, Kenneth L</creator><creator>HAZAN, Elad</creator><creator>WOODRUFF, David P</creator><general>Association for Computing Machinery</general><scope>IQODW</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>7TB</scope><scope>FR3</scope></search><sort><creationdate>20121001</creationdate><title>Sublinear Optimization for Machine Learning</title><author>CLARKSON, Kenneth L ; HAZAN, Elad ; WOODRUFF, David P</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c332t-6d11601e123c8defeb2e00a5e94863cba0d5ff3180bc8e2e52c1a5575b6ddd8e3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2012</creationdate><topic>Algorithmics. Computability. Computer arithmetics</topic><topic>Algorithms</topic><topic>Applied sciences</topic><topic>Artificial intelligence</topic><topic>Branch & bound algorithms</topic><topic>Computer science; control theory; systems</topic><topic>Exact sciences and technology</topic><topic>Linear programming</topic><topic>Machine learning</topic><topic>Optimization algorithms</topic><topic>Studies</topic><topic>Theoretical computing</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>CLARKSON, Kenneth L</creatorcontrib><creatorcontrib>HAZAN, Elad</creatorcontrib><creatorcontrib>WOODRUFF, David P</creatorcontrib><collection>Pascal-Francis</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>Mechanical & Transportation Engineering Abstracts</collection><collection>Engineering Research Database</collection><jtitle>Journal of the ACM</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>CLARKSON, Kenneth L</au><au>HAZAN, Elad</au><au>WOODRUFF, David P</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Sublinear Optimization for Machine Learning</atitle><jtitle>Journal of the ACM</jtitle><date>2012-10-01</date><risdate>2012</risdate><volume>59</volume><issue>5</issue><spage>1</spage><epage>49</epage><pages>1-49</pages><issn>0004-5411</issn><eissn>1557-735X</eissn><coden>JACOAH</coden><abstract>In this article we describe and analyze sublinear-time approximation algorithms for some optimization problems arising in machine learning, such as training linear classifiers and finding minimum enclosing balls. Our algorithms can be extended to some kernelized versions of these problems, such as SVDD, hard margin SVM, and
L
2
-SVM, for which sublinear-time algorithms were not known before. These new algorithms use a combination of a novel sampling techniques and a new multiplicative update algorithm. We give lower bounds which show the running times of many of our algorithms to be nearly best possible in the unit-cost RAM model.</abstract><cop>New York, NY</cop><pub>Association for Computing Machinery</pub><doi>10.1145/2371656.2371658</doi><tpages>49</tpages></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0004-5411 |
| ispartof | Journal of the ACM, 2012-10, Vol.59 (5), p.1-49 |
| issn | 0004-5411 1557-735X |
| language | eng |
| recordid | cdi_proquest_miscellaneous_1506361775 |
| source | ACM Digital Library Complete |
| subjects | Algorithmics. Computability. Computer arithmetics Algorithms Applied sciences Artificial intelligence Branch & bound algorithms Computer science control theory systems Exact sciences and technology Linear programming Machine learning Optimization algorithms Studies Theoretical computing |
| title | Sublinear Optimization for Machine Learning |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-07T00%3A23%3A01IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Sublinear%20Optimization%20for%20Machine%20Learning&rft.jtitle=Journal%20of%20the%20ACM&rft.au=CLARKSON,%20Kenneth%20L&rft.date=2012-10-01&rft.volume=59&rft.issue=5&rft.spage=1&rft.epage=49&rft.pages=1-49&rft.issn=0004-5411&rft.eissn=1557-735X&rft.coden=JACOAH&rft_id=info:doi/10.1145/2371656.2371658&rft_dat=%3Cproquest_cross%3E2818926901%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=1170617617&rft_id=info:pmid/&rfr_iscdi=true |