Partial label learning based on label distributions and error-correcting output codes

Partial label learning (PLL) is a class of weak supervision learning problems in which each data sample has a candidate set of labels, among which only one label is correct. In this paper, a new PLL algorithm with prior information of the label distribution based on ECOC (PL-PIE) is proposed. PL-PIE...

Full description

Saved in:
Bibliographic Details
Published in:Soft computing (Berlin, Germany) Germany), 2021, Vol.25 (2), p.1049-1064
Main Authors: Lin, Guangyi, Liu, Kunhong, Wang, Beizhan, Zhang, Xiaoyan
Format: Article
Language:English
Subjects:
Artificial Intelligence
Computational Intelligence
Control
Engineering
Mathematical Logic and Foundations
Mechatronics
Methodologies and Application
Robotics
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c291t-b77efb5f915f13f79d8ac4d762b899f4481310bb4deeed2bd34a21fb514b9b363
cites cdi_FETCH-LOGICAL-c291t-b77efb5f915f13f79d8ac4d762b899f4481310bb4deeed2bd34a21fb514b9b363
container_end_page 1064
container_issue 2
container_start_page 1049
container_title Soft computing (Berlin, Germany)
container_volume 25
creator Lin, Guangyi
Liu, Kunhong
Wang, Beizhan
Zhang, Xiaoyan
description Partial label learning (PLL) is a class of weak supervision learning problems in which each data sample has a candidate set of labels, among which only one label is correct. In this paper, a new PLL algorithm with prior information of the label distribution based on ECOC (PL-PIE) is proposed. PL-PIE utilizes the ECOC framework to decompose the problem into multiple binary problems. Different from the instability of the existing random dichotomy, the proposal exploits the prior information of label distribution to generate positive and negative classes with stable performance. Extensive experimental results demonstrate that the proposed PL-PIE algorithm has highly competitive performance compared to the state-of-the-art PLL algorithms.
doi_str_mv 10.1007/s00500-020-05203-0
format article
fullrecord <record><control><sourceid>crossref_sprin</sourceid><recordid>TN_cdi_crossref_primary_10_1007_s00500_020_05203_0</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>10_1007_s00500_020_05203_0</sourcerecordid><originalsourceid>FETCH-LOGICAL-c291t-b77efb5f915f13f79d8ac4d762b899f4481310bb4deeed2bd34a21fb514b9b363</originalsourceid><addsrcrecordid>eNp9kMtKQzEQQIMoWKs_4Or-QHTyuE2zlOILCrqw65DcJOWWmpRJ7sK_N21du5gHw5xhOITcM3hgAOqxAPQAFHiLnoOgcEFmTApBlVT68tRzqhZSXJObUnYAnKlezMjm02Id7b7bWxdaDhbTmLadsyX4Lqe_uR9LxdFNdcypdDb5LiBmpENGDEM9Enmqh6l2Q_ah3JKraPcl3P3VOdm8PH-t3uj64_V99bSmA9esUqdUiK6PmvWRiai0X9pBerXgbql1lHLJBAPnpA8heO68kJazRjDptBMLMSf8fHfAXAqGaA44flv8MQzMUYw5izFNjDmJMdAgcYZKW07bgGaXJ0ztz_-oX9R6Z9E</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Partial label learning based on label distributions and error-correcting output codes</title><source>Springer Nature</source><creator>Lin, Guangyi ; Liu, Kunhong ; Wang, Beizhan ; Zhang, Xiaoyan</creator><creatorcontrib>Lin, Guangyi ; Liu, Kunhong ; Wang, Beizhan ; Zhang, Xiaoyan</creatorcontrib><description>Partial label learning (PLL) is a class of weak supervision learning problems in which each data sample has a candidate set of labels, among which only one label is correct. In this paper, a new PLL algorithm with prior information of the label distribution based on ECOC (PL-PIE) is proposed. PL-PIE utilizes the ECOC framework to decompose the problem into multiple binary problems. Different from the instability of the existing random dichotomy, the proposal exploits the prior information of label distribution to generate positive and negative classes with stable performance. Extensive experimental results demonstrate that the proposed PL-PIE algorithm has highly competitive performance compared to the state-of-the-art PLL algorithms.</description><identifier>ISSN: 1432-7643</identifier><identifier>EISSN: 1433-7479</identifier><identifier>DOI: 10.1007/s00500-020-05203-0</identifier><language>eng</language><publisher>Berlin/Heidelberg: Springer Berlin Heidelberg</publisher><subject>Artificial Intelligence ; Computational Intelligence ; Control ; Engineering ; Mathematical Logic and Foundations ; Mechatronics ; Methodologies and Application ; Robotics</subject><ispartof>Soft computing (Berlin, Germany), 2021, Vol.25 (2), p.1049-1064</ispartof><rights>Springer-Verlag GmbH Germany, part of Springer Nature 2020</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c291t-b77efb5f915f13f79d8ac4d762b899f4481310bb4deeed2bd34a21fb514b9b363</citedby><cites>FETCH-LOGICAL-c291t-b77efb5f915f13f79d8ac4d762b899f4481310bb4deeed2bd34a21fb514b9b363</cites><orcidid>0000-0001-7479-5684</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>314,776,780,27901,27902</link.rule.ids></links><search><creatorcontrib>Lin, Guangyi</creatorcontrib><creatorcontrib>Liu, Kunhong</creatorcontrib><creatorcontrib>Wang, Beizhan</creatorcontrib><creatorcontrib>Zhang, Xiaoyan</creatorcontrib><title>Partial label learning based on label distributions and error-correcting output codes</title><title>Soft computing (Berlin, Germany)</title><addtitle>Soft Comput</addtitle><description>Partial label learning (PLL) is a class of weak supervision learning problems in which each data sample has a candidate set of labels, among which only one label is correct. In this paper, a new PLL algorithm with prior information of the label distribution based on ECOC (PL-PIE) is proposed. PL-PIE utilizes the ECOC framework to decompose the problem into multiple binary problems. Different from the instability of the existing random dichotomy, the proposal exploits the prior information of label distribution to generate positive and negative classes with stable performance. Extensive experimental results demonstrate that the proposed PL-PIE algorithm has highly competitive performance compared to the state-of-the-art PLL algorithms.</description><subject>Artificial Intelligence</subject><subject>Computational Intelligence</subject><subject>Control</subject><subject>Engineering</subject><subject>Mathematical Logic and Foundations</subject><subject>Mechatronics</subject><subject>Methodologies and Application</subject><subject>Robotics</subject><issn>1432-7643</issn><issn>1433-7479</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><recordid>eNp9kMtKQzEQQIMoWKs_4Or-QHTyuE2zlOILCrqw65DcJOWWmpRJ7sK_N21du5gHw5xhOITcM3hgAOqxAPQAFHiLnoOgcEFmTApBlVT68tRzqhZSXJObUnYAnKlezMjm02Id7b7bWxdaDhbTmLadsyX4Lqe_uR9LxdFNdcypdDb5LiBmpENGDEM9Enmqh6l2Q_ah3JKraPcl3P3VOdm8PH-t3uj64_V99bSmA9esUqdUiK6PmvWRiai0X9pBerXgbql1lHLJBAPnpA8heO68kJazRjDptBMLMSf8fHfAXAqGaA44flv8MQzMUYw5izFNjDmJMdAgcYZKW07bgGaXJ0ztz_-oX9R6Z9E</recordid><startdate>2021</startdate><enddate>2021</enddate><creator>Lin, Guangyi</creator><creator>Liu, Kunhong</creator><creator>Wang, Beizhan</creator><creator>Zhang, Xiaoyan</creator><general>Springer Berlin Heidelberg</general><scope>AAYXX</scope><scope>CITATION</scope><orcidid>https://orcid.org/0000-0001-7479-5684</orcidid></search><sort><creationdate>2021</creationdate><title>Partial label learning based on label distributions and error-correcting output codes</title><author>Lin, Guangyi ; Liu, Kunhong ; Wang, Beizhan ; Zhang, Xiaoyan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c291t-b77efb5f915f13f79d8ac4d762b899f4481310bb4deeed2bd34a21fb514b9b363</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Artificial Intelligence</topic><topic>Computational Intelligence</topic><topic>Control</topic><topic>Engineering</topic><topic>Mathematical Logic and Foundations</topic><topic>Mechatronics</topic><topic>Methodologies and Application</topic><topic>Robotics</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Lin, Guangyi</creatorcontrib><creatorcontrib>Liu, Kunhong</creatorcontrib><creatorcontrib>Wang, Beizhan</creatorcontrib><creatorcontrib>Zhang, Xiaoyan</creatorcontrib><collection>CrossRef</collection><jtitle>Soft computing (Berlin, Germany)</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Lin, Guangyi</au><au>Liu, Kunhong</au><au>Wang, Beizhan</au><au>Zhang, Xiaoyan</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Partial label learning based on label distributions and error-correcting output codes</atitle><jtitle>Soft computing (Berlin, Germany)</jtitle><stitle>Soft Comput</stitle><date>2021</date><risdate>2021</risdate><volume>25</volume><issue>2</issue><spage>1049</spage><epage>1064</epage><pages>1049-1064</pages><issn>1432-7643</issn><eissn>1433-7479</eissn><abstract>Partial label learning (PLL) is a class of weak supervision learning problems in which each data sample has a candidate set of labels, among which only one label is correct. In this paper, a new PLL algorithm with prior information of the label distribution based on ECOC (PL-PIE) is proposed. PL-PIE utilizes the ECOC framework to decompose the problem into multiple binary problems. Different from the instability of the existing random dichotomy, the proposal exploits the prior information of label distribution to generate positive and negative classes with stable performance. Extensive experimental results demonstrate that the proposed PL-PIE algorithm has highly competitive performance compared to the state-of-the-art PLL algorithms.</abstract><cop>Berlin/Heidelberg</cop><pub>Springer Berlin Heidelberg</pub><doi>10.1007/s00500-020-05203-0</doi><tpages>16</tpages><orcidid>https://orcid.org/0000-0001-7479-5684</orcidid></addata></record>
fulltext fulltext
identifier ISSN: 1432-7643
ispartof Soft computing (Berlin, Germany), 2021, Vol.25 (2), p.1049-1064
issn 1432-7643
1433-7479
language eng
recordid cdi_crossref_primary_10_1007_s00500_020_05203_0
source Springer Nature
subjects Artificial Intelligence
Computational Intelligence
Control
Engineering
Mathematical Logic and Foundations
Mechatronics
Methodologies and Application
Robotics
title Partial label learning based on label distributions and error-correcting output codes
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-30T16%3A01%3A00IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-crossref_sprin&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Partial%20label%20learning%20based%20on%20label%20distributions%20and%20error-correcting%20output%20codes&rft.jtitle=Soft%20computing%20(Berlin,%20Germany)&rft.au=Lin,%20Guangyi&rft.date=2021&rft.volume=25&rft.issue=2&rft.spage=1049&rft.epage=1064&rft.pages=1049-1064&rft.issn=1432-7643&rft.eissn=1433-7479&rft_id=info:doi/10.1007/s00500-020-05203-0&rft_dat=%3Ccrossref_sprin%3E10_1007_s00500_020_05203_0%3C/crossref_sprin%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c291t-b77efb5f915f13f79d8ac4d762b899f4481310bb4deeed2bd34a21fb514b9b363%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true