Robust Maximum Association Estimators

The maximum association between two multivariate variables and is defined as the maximal value that a bivariate association measure between one-dimensional projections and can attain. Taking the Pearson correlation as projection index results in the first canonical correlation coefficient. We propos...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of the American Statistical Association 2017-03, Vol.112 (517), p.436-445
Hauptverfasser: Alfons, Andreas, Croux, Christophe, Filzmoser, Peter
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 445
container_issue 517
container_start_page 436
container_title Journal of the American Statistical Association
container_volume 112
creator Alfons, Andreas
Croux, Christophe
Filzmoser, Peter
description The maximum association between two multivariate variables and is defined as the maximal value that a bivariate association measure between one-dimensional projections and can attain. Taking the Pearson correlation as projection index results in the first canonical correlation coefficient. We propose to use more robust association measures, such as Spearman's or Kendall's rank correlation, or association measures derived from bivariate scatter matrices. We study the robustness of the proposed maximum association measures and the corresponding estimators of the coefficients yielding the maximum association. In the important special case of being univariate, maximum rank correlation estimators yield regression estimators that are invariant against monotonic transformations of the response. We obtain asymptotic variances for this special case. It turns out that maximum rank correlation estimators combine good efficiency and robustness properties. Simulations and a real data example illustrate the robustness and the power for handling nonlinear relationships of these estimators. Supplementary materials for this article are available online.
doi_str_mv 10.1080/01621459.2016.1148609
format Article
fullrecord <record><control><sourceid>jstor_cross</sourceid><recordid>TN_cdi_jstor_primary_45027930</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><jstor_id>45027930</jstor_id><sourcerecordid>45027930</sourcerecordid><originalsourceid>FETCH-LOGICAL-c440t-ca11792b22f5d9a27f9482d79a8124af81ec968199d21019f91e1d224a43b0af3</originalsourceid><addsrcrecordid>eNp9UE1LAzEQDaJgrf6EQkE8bs0k2Sa5WUr9gIogCt5CuruBLd1NzWTR_nuzbPXoXGbgvTdv5hEyAToDqugthTkDkesZS9MMQKg51SdkBDmXGZPi45SMek7Wk87JBeKWppJKjcjNq990GKfP9rtuuma6QPRFbWPt2-kKY93Y6ANekjNnd1hdHfuYvN-v3paP2frl4Wm5WGeFEDRmhQWQmm0Yc3mpLZNOC8VKqa0CJqxTUBV6rkDrkgEF7TRUULIECb6h1vExuR727oP_7CqMZuu70CZLwziXUgqRi8TKB1YRPGKonNmHdGg4GKCmT8T8JmL6RMwxkaSbDLotpqf-RCKnTGpOE3434HXrfGjslw-70kR72Pnggm2LGg3_3-IH0jRvFw</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2337774454</pqid></control><display><type>article</type><title>Robust Maximum Association Estimators</title><source>JSTOR Mathematics and Statistics</source><source>Taylor &amp; Francis Online</source><source>JSTOR</source><creator>Alfons, Andreas ; Croux, Christophe ; Filzmoser, Peter</creator><creatorcontrib>Alfons, Andreas ; Croux, Christophe ; Filzmoser, Peter</creatorcontrib><description>The maximum association between two multivariate variables and is defined as the maximal value that a bivariate association measure between one-dimensional projections and can attain. Taking the Pearson correlation as projection index results in the first canonical correlation coefficient. We propose to use more robust association measures, such as Spearman's or Kendall's rank correlation, or association measures derived from bivariate scatter matrices. We study the robustness of the proposed maximum association measures and the corresponding estimators of the coefficients yielding the maximum association. In the important special case of being univariate, maximum rank correlation estimators yield regression estimators that are invariant against monotonic transformations of the response. We obtain asymptotic variances for this special case. It turns out that maximum rank correlation estimators combine good efficiency and robustness properties. Simulations and a real data example illustrate the robustness and the power for handling nonlinear relationships of these estimators. Supplementary materials for this article are available online.</description><identifier>ISSN: 0162-1459</identifier><identifier>EISSN: 1537-274X</identifier><identifier>DOI: 10.1080/01621459.2016.1148609</identifier><language>eng</language><publisher>Alexandria: Taylor &amp; Francis</publisher><subject>Bivariate analysis ; Correlation analysis ; Correlation coefficients ; Economic models ; Estimators ; Influence function ; Materials handling ; Mathematical analysis ; Matrices ; Power ; Projection pursuit ; Projections ; Rank correlation ; Regression ; Regression analysis ; Robustness ; Statistical methods ; Statistics ; Theory and Methods</subject><ispartof>Journal of the American Statistical Association, 2017-03, Vol.112 (517), p.436-445</ispartof><rights>2017 The Author(s). Published with license by Taylor &amp; Francis © Andreas Alfons, Christophe Croux, and Peter Filzmoser 2017</rights><rights>Copyright © 2017 American Statistical Association</rights><rights>2017 The Author(s). Published with license by Taylor &amp; Francis; © Andreas Alfons, Christophe Croux, and Peter Filzmoser</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c440t-ca11792b22f5d9a27f9482d79a8124af81ec968199d21019f91e1d224a43b0af3</citedby><cites>FETCH-LOGICAL-c440t-ca11792b22f5d9a27f9482d79a8124af81ec968199d21019f91e1d224a43b0af3</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://www.jstor.org/stable/pdf/45027930$$EPDF$$P50$$Gjstor$$H</linktopdf><linktohtml>$$Uhttps://www.jstor.org/stable/45027930$$EHTML$$P50$$Gjstor$$H</linktohtml><link.rule.ids>315,781,785,804,833,27926,27927,58019,58023,58252,58256,59649,60438</link.rule.ids></links><search><creatorcontrib>Alfons, Andreas</creatorcontrib><creatorcontrib>Croux, Christophe</creatorcontrib><creatorcontrib>Filzmoser, Peter</creatorcontrib><title>Robust Maximum Association Estimators</title><title>Journal of the American Statistical Association</title><description>The maximum association between two multivariate variables and is defined as the maximal value that a bivariate association measure between one-dimensional projections and can attain. Taking the Pearson correlation as projection index results in the first canonical correlation coefficient. We propose to use more robust association measures, such as Spearman's or Kendall's rank correlation, or association measures derived from bivariate scatter matrices. We study the robustness of the proposed maximum association measures and the corresponding estimators of the coefficients yielding the maximum association. In the important special case of being univariate, maximum rank correlation estimators yield regression estimators that are invariant against monotonic transformations of the response. We obtain asymptotic variances for this special case. It turns out that maximum rank correlation estimators combine good efficiency and robustness properties. Simulations and a real data example illustrate the robustness and the power for handling nonlinear relationships of these estimators. Supplementary materials for this article are available online.</description><subject>Bivariate analysis</subject><subject>Correlation analysis</subject><subject>Correlation coefficients</subject><subject>Economic models</subject><subject>Estimators</subject><subject>Influence function</subject><subject>Materials handling</subject><subject>Mathematical analysis</subject><subject>Matrices</subject><subject>Power</subject><subject>Projection pursuit</subject><subject>Projections</subject><subject>Rank correlation</subject><subject>Regression</subject><subject>Regression analysis</subject><subject>Robustness</subject><subject>Statistical methods</subject><subject>Statistics</subject><subject>Theory and Methods</subject><issn>0162-1459</issn><issn>1537-274X</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2017</creationdate><recordtype>article</recordtype><sourceid>0YH</sourceid><recordid>eNp9UE1LAzEQDaJgrf6EQkE8bs0k2Sa5WUr9gIogCt5CuruBLd1NzWTR_nuzbPXoXGbgvTdv5hEyAToDqugthTkDkesZS9MMQKg51SdkBDmXGZPi45SMek7Wk87JBeKWppJKjcjNq990GKfP9rtuuma6QPRFbWPt2-kKY93Y6ANekjNnd1hdHfuYvN-v3paP2frl4Wm5WGeFEDRmhQWQmm0Yc3mpLZNOC8VKqa0CJqxTUBV6rkDrkgEF7TRUULIECb6h1vExuR727oP_7CqMZuu70CZLwziXUgqRi8TKB1YRPGKonNmHdGg4GKCmT8T8JmL6RMwxkaSbDLotpqf-RCKnTGpOE3434HXrfGjslw-70kR72Pnggm2LGg3_3-IH0jRvFw</recordid><startdate>20170301</startdate><enddate>20170301</enddate><creator>Alfons, Andreas</creator><creator>Croux, Christophe</creator><creator>Filzmoser, Peter</creator><general>Taylor &amp; Francis</general><general>Taylor &amp; Francis Group,LLC</general><general>Taylor &amp; Francis Ltd</general><scope>0YH</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>8BJ</scope><scope>FQK</scope><scope>JBE</scope><scope>K9.</scope></search><sort><creationdate>20170301</creationdate><title>Robust Maximum Association Estimators</title><author>Alfons, Andreas ; Croux, Christophe ; Filzmoser, Peter</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c440t-ca11792b22f5d9a27f9482d79a8124af81ec968199d21019f91e1d224a43b0af3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2017</creationdate><topic>Bivariate analysis</topic><topic>Correlation analysis</topic><topic>Correlation coefficients</topic><topic>Economic models</topic><topic>Estimators</topic><topic>Influence function</topic><topic>Materials handling</topic><topic>Mathematical analysis</topic><topic>Matrices</topic><topic>Power</topic><topic>Projection pursuit</topic><topic>Projections</topic><topic>Rank correlation</topic><topic>Regression</topic><topic>Regression analysis</topic><topic>Robustness</topic><topic>Statistical methods</topic><topic>Statistics</topic><topic>Theory and Methods</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Alfons, Andreas</creatorcontrib><creatorcontrib>Croux, Christophe</creatorcontrib><creatorcontrib>Filzmoser, Peter</creatorcontrib><collection>Taylor &amp; Francis (Open access)</collection><collection>CrossRef</collection><collection>International Bibliography of the Social Sciences (IBSS)</collection><collection>International Bibliography of the Social Sciences</collection><collection>International Bibliography of the Social Sciences</collection><collection>ProQuest Health &amp; Medical Complete (Alumni)</collection><jtitle>Journal of the American Statistical Association</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Alfons, Andreas</au><au>Croux, Christophe</au><au>Filzmoser, Peter</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Robust Maximum Association Estimators</atitle><jtitle>Journal of the American Statistical Association</jtitle><date>2017-03-01</date><risdate>2017</risdate><volume>112</volume><issue>517</issue><spage>436</spage><epage>445</epage><pages>436-445</pages><issn>0162-1459</issn><eissn>1537-274X</eissn><abstract>The maximum association between two multivariate variables and is defined as the maximal value that a bivariate association measure between one-dimensional projections and can attain. Taking the Pearson correlation as projection index results in the first canonical correlation coefficient. We propose to use more robust association measures, such as Spearman's or Kendall's rank correlation, or association measures derived from bivariate scatter matrices. We study the robustness of the proposed maximum association measures and the corresponding estimators of the coefficients yielding the maximum association. In the important special case of being univariate, maximum rank correlation estimators yield regression estimators that are invariant against monotonic transformations of the response. We obtain asymptotic variances for this special case. It turns out that maximum rank correlation estimators combine good efficiency and robustness properties. Simulations and a real data example illustrate the robustness and the power for handling nonlinear relationships of these estimators. Supplementary materials for this article are available online.</abstract><cop>Alexandria</cop><pub>Taylor &amp; Francis</pub><doi>10.1080/01621459.2016.1148609</doi><tpages>10</tpages><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 0162-1459
ispartof Journal of the American Statistical Association, 2017-03, Vol.112 (517), p.436-445
issn 0162-1459
1537-274X
language eng
recordid cdi_jstor_primary_45027930
source JSTOR Mathematics and Statistics; Taylor & Francis Online; JSTOR
subjects Bivariate analysis
Correlation analysis
Correlation coefficients
Economic models
Estimators
Influence function
Materials handling
Mathematical analysis
Matrices
Power
Projection pursuit
Projections
Rank correlation
Regression
Regression analysis
Robustness
Statistical methods
Statistics
Theory and Methods
title Robust Maximum Association Estimators
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-17T14%3A27%3A01IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-jstor_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Robust%20Maximum%20Association%20Estimators&rft.jtitle=Journal%20of%20the%20American%20Statistical%20Association&rft.au=Alfons,%20Andreas&rft.date=2017-03-01&rft.volume=112&rft.issue=517&rft.spage=436&rft.epage=445&rft.pages=436-445&rft.issn=0162-1459&rft.eissn=1537-274X&rft_id=info:doi/10.1080/01621459.2016.1148609&rft_dat=%3Cjstor_cross%3E45027930%3C/jstor_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2337774454&rft_id=info:pmid/&rft_jstor_id=45027930&rfr_iscdi=true