Decentralized Online Learning Methods Based on Weight-Balancing Over Time-Varying Digraphs
This paper considers decentralized online optimization problems over a graph, where the allocated objective function of each agent is revealed over time and is only known for the corresponding agent in hindsight. Moreover, the graph is directed and time varying. In order to solve the problem, a dece...
Gespeichert in:
| Veröffentlicht in: | IEEE transactions on emerging topics in computational intelligence 2021-06, Vol.5 (3), p.394-406 |
|---|---|
| Hauptverfasser: | , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 406 |
|---|---|
| container_issue | 3 |
| container_start_page | 394 |
| container_title | IEEE transactions on emerging topics in computational intelligence |
| container_volume | 5 |
| creator | Xu, Changqiao Zhu, Junlong Wu, Dapeng Oliver |
| description | This paper considers decentralized online optimization problems over a graph, where the allocated objective function of each agent is revealed over time and is only known for the corresponding agent in hindsight. Moreover, the graph is directed and time varying. In order to solve the problem, a decentralized stochastic subgradient online learning method is proposed over time-varying digraphs. However, the directed graph could generate an asymmetric weight matrix, which is not doubly stochastic matrix. To overcome this difficulty, we employ a weight-balancing technique. By choosing appropriate learning rates, we show that our proposed method can achieve logarithmic regret under strong convexity. Moreover, under convexity, the square-root regret can also be achieved. In addition, numerical simulations in sensor networks for solving the online distributed estimation problem illustrate the theoretical results. |
| doi_str_mv | 10.1109/TETCI.2018.2880771 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_ieee_primary_8543490</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>8543490</ieee_id><sourcerecordid>2532300709</sourcerecordid><originalsourceid>FETCH-LOGICAL-c291t-5a54567e6d13006e481e6c033a23ac3c4462165b4b2f6ab0001e5be8befb56fe3</originalsourceid><addsrcrecordid>eNpNkFtPwkAQhTdGEwnyB_Slic_FvXf7KOCFBMMLXuLLZlumsKRscbeY6K93K8T4NJOZc2ZOPoQuCR4SgvObxd1iPB1STNSQKoWzjJygHuUZSakSb6f_-nM0CGGDMaa5IEzwHnqfQAmu9aa237BM5q62DpIZGO-sWyVP0K6bZUhGJsRt45JXsKt1m45MbVzZKeaf4JOF3UL6YvxXN5nYlTe7dbhAZ5WpAwyOtY-e72PSx3Q2f5iOb2dpSXPSpsIILmQGckkYxhK4IiBLzJihzJSs5FxSIkXBC1pJU8TwBEQBqoCqELIC1kfXh7s733zsIbR60-y9iy81FYzGoxnOo4oeVKVvQvBQ6Z232xhZE6w7jPoXo-4w6iPGaLo6mCwA_BmU4IznmP0AjXtt9A</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2532300709</pqid></control><display><type>article</type><title>Decentralized Online Learning Methods Based on Weight-Balancing Over Time-Varying Digraphs</title><source>IEEE Electronic Library (IEL)</source><creator>Xu, Changqiao ; Zhu, Junlong ; Wu, Dapeng Oliver</creator><creatorcontrib>Xu, Changqiao ; Zhu, Junlong ; Wu, Dapeng Oliver</creatorcontrib><description>This paper considers decentralized online optimization problems over a graph, where the allocated objective function of each agent is revealed over time and is only known for the corresponding agent in hindsight. Moreover, the graph is directed and time varying. In order to solve the problem, a decentralized stochastic subgradient online learning method is proposed over time-varying digraphs. However, the directed graph could generate an asymmetric weight matrix, which is not doubly stochastic matrix. To overcome this difficulty, we employ a weight-balancing technique. By choosing appropriate learning rates, we show that our proposed method can achieve logarithmic regret under strong convexity. Moreover, under convexity, the square-root regret can also be achieved. In addition, numerical simulations in sensor networks for solving the online distributed estimation problem illustrate the theoretical results.</description><identifier>ISSN: 2471-285X</identifier><identifier>EISSN: 2471-285X</identifier><identifier>DOI: 10.1109/TETCI.2018.2880771</identifier><identifier>CODEN: ITETCU</identifier><language>eng</language><publisher>Piscataway: IEEE</publisher><subject>Balancing ; Convex functions ; Convexity ; Cost function ; Decentralized online optimization ; Distance learning ; Estimation ; Graph theory ; Heuristic algorithms ; Learning systems ; Noise measurement ; online algorithm ; Optimization ; regret ; Teaching methods ; time-varying digraphs ; Weight</subject><ispartof>IEEE transactions on emerging topics in computational intelligence, 2021-06, Vol.5 (3), p.394-406</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c291t-5a54567e6d13006e481e6c033a23ac3c4462165b4b2f6ab0001e5be8befb56fe3</citedby><cites>FETCH-LOGICAL-c291t-5a54567e6d13006e481e6c033a23ac3c4462165b4b2f6ab0001e5be8befb56fe3</cites><orcidid>0000-0003-1467-1086 ; 0000-0003-1755-0183</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/8543490$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,777,781,793,27905,27906,54739</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/8543490$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Xu, Changqiao</creatorcontrib><creatorcontrib>Zhu, Junlong</creatorcontrib><creatorcontrib>Wu, Dapeng Oliver</creatorcontrib><title>Decentralized Online Learning Methods Based on Weight-Balancing Over Time-Varying Digraphs</title><title>IEEE transactions on emerging topics in computational intelligence</title><addtitle>TETCI</addtitle><description>This paper considers decentralized online optimization problems over a graph, where the allocated objective function of each agent is revealed over time and is only known for the corresponding agent in hindsight. Moreover, the graph is directed and time varying. In order to solve the problem, a decentralized stochastic subgradient online learning method is proposed over time-varying digraphs. However, the directed graph could generate an asymmetric weight matrix, which is not doubly stochastic matrix. To overcome this difficulty, we employ a weight-balancing technique. By choosing appropriate learning rates, we show that our proposed method can achieve logarithmic regret under strong convexity. Moreover, under convexity, the square-root regret can also be achieved. In addition, numerical simulations in sensor networks for solving the online distributed estimation problem illustrate the theoretical results.</description><subject>Balancing</subject><subject>Convex functions</subject><subject>Convexity</subject><subject>Cost function</subject><subject>Decentralized online optimization</subject><subject>Distance learning</subject><subject>Estimation</subject><subject>Graph theory</subject><subject>Heuristic algorithms</subject><subject>Learning systems</subject><subject>Noise measurement</subject><subject>online algorithm</subject><subject>Optimization</subject><subject>regret</subject><subject>Teaching methods</subject><subject>time-varying digraphs</subject><subject>Weight</subject><issn>2471-285X</issn><issn>2471-285X</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNpNkFtPwkAQhTdGEwnyB_Slic_FvXf7KOCFBMMLXuLLZlumsKRscbeY6K93K8T4NJOZc2ZOPoQuCR4SgvObxd1iPB1STNSQKoWzjJygHuUZSakSb6f_-nM0CGGDMaa5IEzwHnqfQAmu9aa237BM5q62DpIZGO-sWyVP0K6bZUhGJsRt45JXsKt1m45MbVzZKeaf4JOF3UL6YvxXN5nYlTe7dbhAZ5WpAwyOtY-e72PSx3Q2f5iOb2dpSXPSpsIILmQGckkYxhK4IiBLzJihzJSs5FxSIkXBC1pJU8TwBEQBqoCqELIC1kfXh7s733zsIbR60-y9iy81FYzGoxnOo4oeVKVvQvBQ6Z232xhZE6w7jPoXo-4w6iPGaLo6mCwA_BmU4IznmP0AjXtt9A</recordid><startdate>20210601</startdate><enddate>20210601</enddate><creator>Xu, Changqiao</creator><creator>Zhu, Junlong</creator><creator>Wu, Dapeng Oliver</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SP</scope><scope>8FD</scope><scope>L7M</scope><orcidid>https://orcid.org/0000-0003-1467-1086</orcidid><orcidid>https://orcid.org/0000-0003-1755-0183</orcidid></search><sort><creationdate>20210601</creationdate><title>Decentralized Online Learning Methods Based on Weight-Balancing Over Time-Varying Digraphs</title><author>Xu, Changqiao ; Zhu, Junlong ; Wu, Dapeng Oliver</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c291t-5a54567e6d13006e481e6c033a23ac3c4462165b4b2f6ab0001e5be8befb56fe3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Balancing</topic><topic>Convex functions</topic><topic>Convexity</topic><topic>Cost function</topic><topic>Decentralized online optimization</topic><topic>Distance learning</topic><topic>Estimation</topic><topic>Graph theory</topic><topic>Heuristic algorithms</topic><topic>Learning systems</topic><topic>Noise measurement</topic><topic>online algorithm</topic><topic>Optimization</topic><topic>regret</topic><topic>Teaching methods</topic><topic>time-varying digraphs</topic><topic>Weight</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Xu, Changqiao</creatorcontrib><creatorcontrib>Zhu, Junlong</creatorcontrib><creatorcontrib>Wu, Dapeng Oliver</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>CrossRef</collection><collection>Electronics & Communications Abstracts</collection><collection>Technology Research Database</collection><collection>Advanced Technologies Database with Aerospace</collection><jtitle>IEEE transactions on emerging topics in computational intelligence</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Xu, Changqiao</au><au>Zhu, Junlong</au><au>Wu, Dapeng Oliver</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Decentralized Online Learning Methods Based on Weight-Balancing Over Time-Varying Digraphs</atitle><jtitle>IEEE transactions on emerging topics in computational intelligence</jtitle><stitle>TETCI</stitle><date>2021-06-01</date><risdate>2021</risdate><volume>5</volume><issue>3</issue><spage>394</spage><epage>406</epage><pages>394-406</pages><issn>2471-285X</issn><eissn>2471-285X</eissn><coden>ITETCU</coden><abstract>This paper considers decentralized online optimization problems over a graph, where the allocated objective function of each agent is revealed over time and is only known for the corresponding agent in hindsight. Moreover, the graph is directed and time varying. In order to solve the problem, a decentralized stochastic subgradient online learning method is proposed over time-varying digraphs. However, the directed graph could generate an asymmetric weight matrix, which is not doubly stochastic matrix. To overcome this difficulty, we employ a weight-balancing technique. By choosing appropriate learning rates, we show that our proposed method can achieve logarithmic regret under strong convexity. Moreover, under convexity, the square-root regret can also be achieved. In addition, numerical simulations in sensor networks for solving the online distributed estimation problem illustrate the theoretical results.</abstract><cop>Piscataway</cop><pub>IEEE</pub><doi>10.1109/TETCI.2018.2880771</doi><tpages>13</tpages><orcidid>https://orcid.org/0000-0003-1467-1086</orcidid><orcidid>https://orcid.org/0000-0003-1755-0183</orcidid></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 2471-285X |
| ispartof | IEEE transactions on emerging topics in computational intelligence, 2021-06, Vol.5 (3), p.394-406 |
| issn | 2471-285X 2471-285X |
| language | eng |
| recordid | cdi_ieee_primary_8543490 |
| source | IEEE Electronic Library (IEL) |
| subjects | Balancing Convex functions Convexity Cost function Decentralized online optimization Distance learning Estimation Graph theory Heuristic algorithms Learning systems Noise measurement online algorithm Optimization regret Teaching methods time-varying digraphs Weight |
| title | Decentralized Online Learning Methods Based on Weight-Balancing Over Time-Varying Digraphs |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-19T10%3A08%3A25IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Decentralized%20Online%20Learning%20Methods%20Based%20on%20Weight-Balancing%20Over%20Time-Varying%20Digraphs&rft.jtitle=IEEE%20transactions%20on%20emerging%20topics%20in%20computational%20intelligence&rft.au=Xu,%20Changqiao&rft.date=2021-06-01&rft.volume=5&rft.issue=3&rft.spage=394&rft.epage=406&rft.pages=394-406&rft.issn=2471-285X&rft.eissn=2471-285X&rft.coden=ITETCU&rft_id=info:doi/10.1109/TETCI.2018.2880771&rft_dat=%3Cproquest_RIE%3E2532300709%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2532300709&rft_id=info:pmid/&rft_ieee_id=8543490&rfr_iscdi=true |