Maximum Independent Component Analysis with Application to EEG Data

In many scientific disciplines, finding hidden influential factors behind observational data is essential but challenging. The majority of existing approaches, such as the independent component analysis (ICA), rely on linear transformation, that is, true signals are linear combinations of hidden com...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Statistical science 2020-02, Vol.35 (1), p.145-157
Hauptverfasser: Guo, Ruosi, Zhang, Chunming, Zhang, Zhengjun
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 157
container_issue 1
container_start_page 145
container_title Statistical science
container_volume 35
creator Guo, Ruosi
Zhang, Chunming
Zhang, Zhengjun
description In many scientific disciplines, finding hidden influential factors behind observational data is essential but challenging. The majority of existing approaches, such as the independent component analysis (ICA), rely on linear transformation, that is, true signals are linear combinations of hidden components. Motivated from analyzing nonlinear temporal signals in neuroscience, genetics, and finance, this paper proposes the "maximum independent component analysis" (MaxICA), based on max-linear combinations of components. In contrast to existing methods, MaxICA benefits from focusing on significant major components while filtering out ignorable components. A major tool for parameter learning of MaxICA is an augmented genetic algorithm, consisting of three schemes for the elite weighted sum selection, randomly combined crossover, and dynamic mutation. Extensive empirical evaluations demonstrate the effectiveness of MaxICA in either extracting max-linearly combined essential sources in many applications or supplying a better approximation for nonlinearly combined source signals, such as EEG recordings analyzed in this paper.
doi_str_mv 10.1214/19-STS763
format Article
fullrecord <record><control><sourceid>jstor_proqu</sourceid><recordid>TN_cdi_proquest_journals_2384187062</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><jstor_id>26997883</jstor_id><sourcerecordid>26997883</sourcerecordid><originalsourceid>FETCH-LOGICAL-c314t-84b80a609aea0e54a91b98368a5fd375783836fbe28708203bea683bb0b3c00f3</originalsourceid><addsrcrecordid>eNo9kE1LAzEQhoMoWKsHf4AQ8ORhNV-bj2NZay1UPLSeQ7LN4pbuZk1StP_elBUvM_PCM8PwAHCL0SMmmD1hVaw3a8HpGZgQzGUhBSvPwQRJSQtGqLgEVzHuEEIlx2wCqjfz03aHDi77rRtcLn2Cle8G35-mWW_2x9hG-N2mTzgbhn1bm9T6HiYP5_MFfDbJXIOLxuyju_nrU_DxMt9Ur8XqfbGsZquippilQjIrkeFIGWeQK5lR2CpJuTRls6WiFJLm1FhHpECSIGqd4ZJaiyytEWroFNyPd4fgvw4uJr3zh5A_jJpQyXBe4yRTDyNVBx9jcI0eQtuZcNQY6ZMjjZUeHWX2bmR3MfnwDxKulMjC6C_a3WE6</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2384187062</pqid></control><display><type>article</type><title>Maximum Independent Component Analysis with Application to EEG Data</title><source>JSTOR Mathematics &amp; Statistics</source><source>JSTOR Archive Collection A-Z Listing</source><source>EZB-FREE-00999 freely available EZB journals</source><source>Project Euclid Complete</source><creator>Guo, Ruosi ; Zhang, Chunming ; Zhang, Zhengjun</creator><creatorcontrib>Guo, Ruosi ; Zhang, Chunming ; Zhang, Zhengjun</creatorcontrib><description>In many scientific disciplines, finding hidden influential factors behind observational data is essential but challenging. The majority of existing approaches, such as the independent component analysis (ICA), rely on linear transformation, that is, true signals are linear combinations of hidden components. Motivated from analyzing nonlinear temporal signals in neuroscience, genetics, and finance, this paper proposes the "maximum independent component analysis" (MaxICA), based on max-linear combinations of components. In contrast to existing methods, MaxICA benefits from focusing on significant major components while filtering out ignorable components. A major tool for parameter learning of MaxICA is an augmented genetic algorithm, consisting of three schemes for the elite weighted sum selection, randomly combined crossover, and dynamic mutation. Extensive empirical evaluations demonstrate the effectiveness of MaxICA in either extracting max-linearly combined essential sources in many applications or supplying a better approximation for nonlinearly combined source signals, such as EEG recordings analyzed in this paper.</description><identifier>ISSN: 0883-4237</identifier><identifier>EISSN: 2168-8745</identifier><identifier>DOI: 10.1214/19-STS763</identifier><language>eng</language><publisher>Hayward: Institute of Mathematical Statistics</publisher><subject>Crossovers ; Data analysis ; Electroencephalography ; Empirical analysis ; Genetic algorithms ; Independent component analysis ; Linear transformations ; Machine learning ; Nonlinear analysis ; Principal components analysis</subject><ispartof>Statistical science, 2020-02, Vol.35 (1), p.145-157</ispartof><rights>Institute of Mathematical Statistics, 2020</rights><rights>Copyright Institute of Mathematical Statistics Feb 2020</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c314t-84b80a609aea0e54a91b98368a5fd375783836fbe28708203bea683bb0b3c00f3</citedby><cites>FETCH-LOGICAL-c314t-84b80a609aea0e54a91b98368a5fd375783836fbe28708203bea683bb0b3c00f3</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://www.jstor.org/stable/pdf/26997883$$EPDF$$P50$$Gjstor$$H</linktopdf><linktohtml>$$Uhttps://www.jstor.org/stable/26997883$$EHTML$$P50$$Gjstor$$H</linktohtml><link.rule.ids>314,780,784,803,832,27924,27925,58017,58021,58250,58254</link.rule.ids></links><search><creatorcontrib>Guo, Ruosi</creatorcontrib><creatorcontrib>Zhang, Chunming</creatorcontrib><creatorcontrib>Zhang, Zhengjun</creatorcontrib><title>Maximum Independent Component Analysis with Application to EEG Data</title><title>Statistical science</title><description>In many scientific disciplines, finding hidden influential factors behind observational data is essential but challenging. The majority of existing approaches, such as the independent component analysis (ICA), rely on linear transformation, that is, true signals are linear combinations of hidden components. Motivated from analyzing nonlinear temporal signals in neuroscience, genetics, and finance, this paper proposes the "maximum independent component analysis" (MaxICA), based on max-linear combinations of components. In contrast to existing methods, MaxICA benefits from focusing on significant major components while filtering out ignorable components. A major tool for parameter learning of MaxICA is an augmented genetic algorithm, consisting of three schemes for the elite weighted sum selection, randomly combined crossover, and dynamic mutation. Extensive empirical evaluations demonstrate the effectiveness of MaxICA in either extracting max-linearly combined essential sources in many applications or supplying a better approximation for nonlinearly combined source signals, such as EEG recordings analyzed in this paper.</description><subject>Crossovers</subject><subject>Data analysis</subject><subject>Electroencephalography</subject><subject>Empirical analysis</subject><subject>Genetic algorithms</subject><subject>Independent component analysis</subject><subject>Linear transformations</subject><subject>Machine learning</subject><subject>Nonlinear analysis</subject><subject>Principal components analysis</subject><issn>0883-4237</issn><issn>2168-8745</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><recordid>eNo9kE1LAzEQhoMoWKsHf4AQ8ORhNV-bj2NZay1UPLSeQ7LN4pbuZk1StP_elBUvM_PCM8PwAHCL0SMmmD1hVaw3a8HpGZgQzGUhBSvPwQRJSQtGqLgEVzHuEEIlx2wCqjfz03aHDi77rRtcLn2Cle8G35-mWW_2x9hG-N2mTzgbhn1bm9T6HiYP5_MFfDbJXIOLxuyju_nrU_DxMt9Ur8XqfbGsZquippilQjIrkeFIGWeQK5lR2CpJuTRls6WiFJLm1FhHpECSIGqd4ZJaiyytEWroFNyPd4fgvw4uJr3zh5A_jJpQyXBe4yRTDyNVBx9jcI0eQtuZcNQY6ZMjjZUeHWX2bmR3MfnwDxKulMjC6C_a3WE6</recordid><startdate>20200201</startdate><enddate>20200201</enddate><creator>Guo, Ruosi</creator><creator>Zhang, Chunming</creator><creator>Zhang, Zhengjun</creator><general>Institute of Mathematical Statistics</general><scope>AAYXX</scope><scope>CITATION</scope></search><sort><creationdate>20200201</creationdate><title>Maximum Independent Component Analysis with Application to EEG Data</title><author>Guo, Ruosi ; Zhang, Chunming ; Zhang, Zhengjun</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c314t-84b80a609aea0e54a91b98368a5fd375783836fbe28708203bea683bb0b3c00f3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Crossovers</topic><topic>Data analysis</topic><topic>Electroencephalography</topic><topic>Empirical analysis</topic><topic>Genetic algorithms</topic><topic>Independent component analysis</topic><topic>Linear transformations</topic><topic>Machine learning</topic><topic>Nonlinear analysis</topic><topic>Principal components analysis</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Guo, Ruosi</creatorcontrib><creatorcontrib>Zhang, Chunming</creatorcontrib><creatorcontrib>Zhang, Zhengjun</creatorcontrib><collection>CrossRef</collection><jtitle>Statistical science</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Guo, Ruosi</au><au>Zhang, Chunming</au><au>Zhang, Zhengjun</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Maximum Independent Component Analysis with Application to EEG Data</atitle><jtitle>Statistical science</jtitle><date>2020-02-01</date><risdate>2020</risdate><volume>35</volume><issue>1</issue><spage>145</spage><epage>157</epage><pages>145-157</pages><issn>0883-4237</issn><eissn>2168-8745</eissn><abstract>In many scientific disciplines, finding hidden influential factors behind observational data is essential but challenging. The majority of existing approaches, such as the independent component analysis (ICA), rely on linear transformation, that is, true signals are linear combinations of hidden components. Motivated from analyzing nonlinear temporal signals in neuroscience, genetics, and finance, this paper proposes the "maximum independent component analysis" (MaxICA), based on max-linear combinations of components. In contrast to existing methods, MaxICA benefits from focusing on significant major components while filtering out ignorable components. A major tool for parameter learning of MaxICA is an augmented genetic algorithm, consisting of three schemes for the elite weighted sum selection, randomly combined crossover, and dynamic mutation. Extensive empirical evaluations demonstrate the effectiveness of MaxICA in either extracting max-linearly combined essential sources in many applications or supplying a better approximation for nonlinearly combined source signals, such as EEG recordings analyzed in this paper.</abstract><cop>Hayward</cop><pub>Institute of Mathematical Statistics</pub><doi>10.1214/19-STS763</doi><tpages>13</tpages><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 0883-4237
ispartof Statistical science, 2020-02, Vol.35 (1), p.145-157
issn 0883-4237
2168-8745
language eng
recordid cdi_proquest_journals_2384187062
source JSTOR Mathematics & Statistics; JSTOR Archive Collection A-Z Listing; EZB-FREE-00999 freely available EZB journals; Project Euclid Complete
subjects Crossovers
Data analysis
Electroencephalography
Empirical analysis
Genetic algorithms
Independent component analysis
Linear transformations
Machine learning
Nonlinear analysis
Principal components analysis
title Maximum Independent Component Analysis with Application to EEG Data
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-06T21%3A16%3A14IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-jstor_proqu&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Maximum%20Independent%20Component%20Analysis%20with%20Application%20to%20EEG%20Data&rft.jtitle=Statistical%20science&rft.au=Guo,%20Ruosi&rft.date=2020-02-01&rft.volume=35&rft.issue=1&rft.spage=145&rft.epage=157&rft.pages=145-157&rft.issn=0883-4237&rft.eissn=2168-8745&rft_id=info:doi/10.1214/19-STS763&rft_dat=%3Cjstor_proqu%3E26997883%3C/jstor_proqu%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2384187062&rft_id=info:pmid/&rft_jstor_id=26997883&rfr_iscdi=true