Learning to rank for censored survival data

Survival analysis is a type of semi-supervised ranking task where the target output (the survival time) is often right-censored. Utilizing this information is a challenge because it is not obvious how to correctly incorporate these censored examples into a model. We study how three categories of los...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2018-06
Main Authors: Luck, Margaux, Sylvain, Tristan, Cohen, Joseph Paul, Cardinal, Heloise, Lodi, Andrea, Bengio, Yoshua
Format: Article
Language:English
Subjects:
Ranking
Survival
Survival analysis
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Luck, Margaux
Sylvain, Tristan
Cohen, Joseph Paul
Cardinal, Heloise
Lodi, Andrea
Bengio, Yoshua
description Survival analysis is a type of semi-supervised ranking task where the target output (the survival time) is often right-censored. Utilizing this information is a challenge because it is not obvious how to correctly incorporate these censored examples into a model. We study how three categories of loss functions, namely partial likelihood methods, rank methods, and our classification method based on a Wasserstein metric (WM) and the non-parametric Kaplan Meier estimate of the probability density to impute the labels of censored examples, can take advantage of this information. The proposed method allows us to have a model that predict the probability distribution of an event. If a clinician had access to the detailed probability of an event over time this would help in treatment planning. For example, determining if the risk of kidney graft rejection is constant or peaked after some time. Also, we demonstrate that this approach directly optimizes the expected C-index which is the most common evaluation metric for ranking survival models.
format article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2073368080</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2073368080</sourcerecordid><originalsourceid>FETCH-proquest_journals_20733680803</originalsourceid><addsrcrecordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mTQ9klNLMrLzEtXKMlXKErMy1ZIyy9SSE7NK84vSk1RKC4tKsssS8xRSEksSeRhYE1LzClO5YXS3AzKbq4hzh66BUX5haWpxSXxWfmlRXlAqXgjA3NjYzMLAwsDY-JUAQBFSzEp</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2073368080</pqid></control><display><type>article</type><title>Learning to rank for censored survival data</title><source>ProQuest - Publicly Available Content Database</source><creator>Luck, Margaux ; Sylvain, Tristan ; Cohen, Joseph Paul ; Cardinal, Heloise ; Lodi, Andrea ; Bengio, Yoshua</creator><creatorcontrib>Luck, Margaux ; Sylvain, Tristan ; Cohen, Joseph Paul ; Cardinal, Heloise ; Lodi, Andrea ; Bengio, Yoshua</creatorcontrib><description>Survival analysis is a type of semi-supervised ranking task where the target output (the survival time) is often right-censored. Utilizing this information is a challenge because it is not obvious how to correctly incorporate these censored examples into a model. We study how three categories of loss functions, namely partial likelihood methods, rank methods, and our classification method based on a Wasserstein metric (WM) and the non-parametric Kaplan Meier estimate of the probability density to impute the labels of censored examples, can take advantage of this information. The proposed method allows us to have a model that predict the probability distribution of an event. If a clinician had access to the detailed probability of an event over time this would help in treatment planning. For example, determining if the risk of kidney graft rejection is constant or peaked after some time. Also, we demonstrate that this approach directly optimizes the expected C-index which is the most common evaluation metric for ranking survival models.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Ranking ; Survival ; Survival analysis</subject><ispartof>arXiv.org, 2018-06</ispartof><rights>2018. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://www.proquest.com/docview/2073368080?pq-origsite=primo$$EHTML$$P50$$Gproquest$$Hfree_for_read</linktohtml><link.rule.ids>780,784,25753,37012,44590</link.rule.ids></links><search><creatorcontrib>Luck, Margaux</creatorcontrib><creatorcontrib>Sylvain, Tristan</creatorcontrib><creatorcontrib>Cohen, Joseph Paul</creatorcontrib><creatorcontrib>Cardinal, Heloise</creatorcontrib><creatorcontrib>Lodi, Andrea</creatorcontrib><creatorcontrib>Bengio, Yoshua</creatorcontrib><title>Learning to rank for censored survival data</title><title>arXiv.org</title><description>Survival analysis is a type of semi-supervised ranking task where the target output (the survival time) is often right-censored. Utilizing this information is a challenge because it is not obvious how to correctly incorporate these censored examples into a model. We study how three categories of loss functions, namely partial likelihood methods, rank methods, and our classification method based on a Wasserstein metric (WM) and the non-parametric Kaplan Meier estimate of the probability density to impute the labels of censored examples, can take advantage of this information. The proposed method allows us to have a model that predict the probability distribution of an event. If a clinician had access to the detailed probability of an event over time this would help in treatment planning. For example, determining if the risk of kidney graft rejection is constant or peaked after some time. Also, we demonstrate that this approach directly optimizes the expected C-index which is the most common evaluation metric for ranking survival models.</description><subject>Ranking</subject><subject>Survival</subject><subject>Survival analysis</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2018</creationdate><recordtype>article</recordtype><sourceid>PIMPY</sourceid><recordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mTQ9klNLMrLzEtXKMlXKErMy1ZIyy9SSE7NK84vSk1RKC4tKsssS8xRSEksSeRhYE1LzClO5YXS3AzKbq4hzh66BUX5haWpxSXxWfmlRXlAqXgjA3NjYzMLAwsDY-JUAQBFSzEp</recordid><startdate>20180608</startdate><enddate>20180608</enddate><creator>Luck, Margaux</creator><creator>Sylvain, Tristan</creator><creator>Cohen, Joseph Paul</creator><creator>Cardinal, Heloise</creator><creator>Lodi, Andrea</creator><creator>Bengio, Yoshua</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20180608</creationdate><title>Learning to rank for censored survival data</title><author>Luck, Margaux ; Sylvain, Tristan ; Cohen, Joseph Paul ; Cardinal, Heloise ; Lodi, Andrea ; Bengio, Yoshua</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_20733680803</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2018</creationdate><topic>Ranking</topic><topic>Survival</topic><topic>Survival analysis</topic><toplevel>online_resources</toplevel><creatorcontrib>Luck, Margaux</creatorcontrib><creatorcontrib>Sylvain, Tristan</creatorcontrib><creatorcontrib>Cohen, Joseph Paul</creatorcontrib><creatorcontrib>Cardinal, Heloise</creatorcontrib><creatorcontrib>Lodi, Andrea</creatorcontrib><creatorcontrib>Bengio, Yoshua</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni)</collection><collection>ProQuest Central</collection><collection>ProQuest Central Essentials</collection><collection>AUTh Library subscriptions: ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>ProQuest - Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Luck, Margaux</au><au>Sylvain, Tristan</au><au>Cohen, Joseph Paul</au><au>Cardinal, Heloise</au><au>Lodi, Andrea</au><au>Bengio, Yoshua</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Learning to rank for censored survival data</atitle><jtitle>arXiv.org</jtitle><date>2018-06-08</date><risdate>2018</risdate><eissn>2331-8422</eissn><abstract>Survival analysis is a type of semi-supervised ranking task where the target output (the survival time) is often right-censored. Utilizing this information is a challenge because it is not obvious how to correctly incorporate these censored examples into a model. We study how three categories of loss functions, namely partial likelihood methods, rank methods, and our classification method based on a Wasserstein metric (WM) and the non-parametric Kaplan Meier estimate of the probability density to impute the labels of censored examples, can take advantage of this information. The proposed method allows us to have a model that predict the probability distribution of an event. If a clinician had access to the detailed probability of an event over time this would help in treatment planning. For example, determining if the risk of kidney graft rejection is constant or peaked after some time. Also, we demonstrate that this approach directly optimizes the expected C-index which is the most common evaluation metric for ranking survival models.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2018-06
issn 2331-8422
language eng
recordid cdi_proquest_journals_2073368080
source ProQuest - Publicly Available Content Database
subjects Ranking
Survival
Survival analysis
title Learning to rank for censored survival data
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-02T09%3A05%3A52IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Learning%20to%20rank%20for%20censored%20survival%20data&rft.jtitle=arXiv.org&rft.au=Luck,%20Margaux&rft.date=2018-06-08&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2073368080%3C/proquest%3E%3Cgrp_id%3Ecdi_FETCH-proquest_journals_20733680803%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2073368080&rft_id=info:pmid/&rfr_iscdi=true