Statistical Inference for Polyak-Ruppert Averaged Zeroth-order Stochastic Gradient Algorithm

Statistical machine learning models trained with stochastic gradient algorithms are increasingly being deployed in critical scientific applications. However, computing the stochastic gradient in several such applications is highly expensive or even impossible at times. In such cases, derivative-free...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Jin, Yanhao, Xiao, Tesi, Balasubramanian, Krishnakumar
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Jin, Yanhao
Xiao, Tesi
Balasubramanian, Krishnakumar
description Statistical machine learning models trained with stochastic gradient algorithms are increasingly being deployed in critical scientific applications. However, computing the stochastic gradient in several such applications is highly expensive or even impossible at times. In such cases, derivative-free or zeroth-order algorithms are used. An important question which has thus far not been addressed sufficiently in the statistical machine learning literature is that of equipping stochastic zeroth-order algorithms with practical yet rigorous inferential capabilities so that we not only have point estimates or predictions but also quantify the associated uncertainty via confidence intervals or sets. Towards this, in this work, we first establish a central limit theorem for Polyak-Ruppert averaged stochastic zeroth-order gradient algorithm. We then provide online estimators of the asymptotic covariance matrix appearing in the central limit theorem, thereby providing a practical procedure for constructing asymptotically valid confidence sets (or intervals) for parameter estimation (or prediction) in the zeroth-order setting.
doi_str_mv 10.48550/arxiv.2102.05198
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2102_05198</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2102_05198</sourcerecordid><originalsourceid>FETCH-LOGICAL-a678-93fbbd8e889d0505d99c733a64305e39d174b20521a6674e802c9050fa5b48a13</originalsourceid><addsrcrecordid>eNotj7FuwjAURb10qKAf0Kn-gaR2HCf2iFBLkZCKClOFFL3EL8RqiKOHi8rfF2inu5x7pMPYoxRpbrQWz0A__pRmUmSp0NKae7bbRIj-GH0DPV8OLRIODfI2EF-H_gxfycf3OCJFPjshwR4d_0QKsUsCOSS-iaHp4PrnCwLncbiQ_T6Qj91hyu5a6I_48L8Ttn192c7fktX7YjmfrRIoSpNY1da1M2iMdUIL7axtSqWgyJXQqKyTZV5nQmcSiqLM0YissRewBV3nBqSasKc_7S2vGskfgM7VNbO6ZapfscRNrQ</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Statistical Inference for Polyak-Ruppert Averaged Zeroth-order Stochastic Gradient Algorithm</title><source>arXiv.org</source><creator>Jin, Yanhao ; Xiao, Tesi ; Balasubramanian, Krishnakumar</creator><creatorcontrib>Jin, Yanhao ; Xiao, Tesi ; Balasubramanian, Krishnakumar</creatorcontrib><description>Statistical machine learning models trained with stochastic gradient algorithms are increasingly being deployed in critical scientific applications. However, computing the stochastic gradient in several such applications is highly expensive or even impossible at times. In such cases, derivative-free or zeroth-order algorithms are used. An important question which has thus far not been addressed sufficiently in the statistical machine learning literature is that of equipping stochastic zeroth-order algorithms with practical yet rigorous inferential capabilities so that we not only have point estimates or predictions but also quantify the associated uncertainty via confidence intervals or sets. Towards this, in this work, we first establish a central limit theorem for Polyak-Ruppert averaged stochastic zeroth-order gradient algorithm. We then provide online estimators of the asymptotic covariance matrix appearing in the central limit theorem, thereby providing a practical procedure for constructing asymptotically valid confidence sets (or intervals) for parameter estimation (or prediction) in the zeroth-order setting.</description><identifier>DOI: 10.48550/arxiv.2102.05198</identifier><language>eng</language><subject>Computer Science - Learning ; Statistics - Machine Learning</subject><creationdate>2021-02</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2102.05198$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2102.05198$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Jin, Yanhao</creatorcontrib><creatorcontrib>Xiao, Tesi</creatorcontrib><creatorcontrib>Balasubramanian, Krishnakumar</creatorcontrib><title>Statistical Inference for Polyak-Ruppert Averaged Zeroth-order Stochastic Gradient Algorithm</title><description>Statistical machine learning models trained with stochastic gradient algorithms are increasingly being deployed in critical scientific applications. However, computing the stochastic gradient in several such applications is highly expensive or even impossible at times. In such cases, derivative-free or zeroth-order algorithms are used. An important question which has thus far not been addressed sufficiently in the statistical machine learning literature is that of equipping stochastic zeroth-order algorithms with practical yet rigorous inferential capabilities so that we not only have point estimates or predictions but also quantify the associated uncertainty via confidence intervals or sets. Towards this, in this work, we first establish a central limit theorem for Polyak-Ruppert averaged stochastic zeroth-order gradient algorithm. We then provide online estimators of the asymptotic covariance matrix appearing in the central limit theorem, thereby providing a practical procedure for constructing asymptotically valid confidence sets (or intervals) for parameter estimation (or prediction) in the zeroth-order setting.</description><subject>Computer Science - Learning</subject><subject>Statistics - Machine Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj7FuwjAURb10qKAf0Kn-gaR2HCf2iFBLkZCKClOFFL3EL8RqiKOHi8rfF2inu5x7pMPYoxRpbrQWz0A__pRmUmSp0NKae7bbRIj-GH0DPV8OLRIODfI2EF-H_gxfycf3OCJFPjshwR4d_0QKsUsCOSS-iaHp4PrnCwLncbiQ_T6Qj91hyu5a6I_48L8Ttn192c7fktX7YjmfrRIoSpNY1da1M2iMdUIL7axtSqWgyJXQqKyTZV5nQmcSiqLM0YissRewBV3nBqSasKc_7S2vGskfgM7VNbO6ZapfscRNrQ</recordid><startdate>20210209</startdate><enddate>20210209</enddate><creator>Jin, Yanhao</creator><creator>Xiao, Tesi</creator><creator>Balasubramanian, Krishnakumar</creator><scope>AKY</scope><scope>EPD</scope><scope>GOX</scope></search><sort><creationdate>20210209</creationdate><title>Statistical Inference for Polyak-Ruppert Averaged Zeroth-order Stochastic Gradient Algorithm</title><author>Jin, Yanhao ; Xiao, Tesi ; Balasubramanian, Krishnakumar</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a678-93fbbd8e889d0505d99c733a64305e39d174b20521a6674e802c9050fa5b48a13</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Computer Science - Learning</topic><topic>Statistics - Machine Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Jin, Yanhao</creatorcontrib><creatorcontrib>Xiao, Tesi</creatorcontrib><creatorcontrib>Balasubramanian, Krishnakumar</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv Statistics</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Jin, Yanhao</au><au>Xiao, Tesi</au><au>Balasubramanian, Krishnakumar</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Statistical Inference for Polyak-Ruppert Averaged Zeroth-order Stochastic Gradient Algorithm</atitle><date>2021-02-09</date><risdate>2021</risdate><abstract>Statistical machine learning models trained with stochastic gradient algorithms are increasingly being deployed in critical scientific applications. However, computing the stochastic gradient in several such applications is highly expensive or even impossible at times. In such cases, derivative-free or zeroth-order algorithms are used. An important question which has thus far not been addressed sufficiently in the statistical machine learning literature is that of equipping stochastic zeroth-order algorithms with practical yet rigorous inferential capabilities so that we not only have point estimates or predictions but also quantify the associated uncertainty via confidence intervals or sets. Towards this, in this work, we first establish a central limit theorem for Polyak-Ruppert averaged stochastic zeroth-order gradient algorithm. We then provide online estimators of the asymptotic covariance matrix appearing in the central limit theorem, thereby providing a practical procedure for constructing asymptotically valid confidence sets (or intervals) for parameter estimation (or prediction) in the zeroth-order setting.</abstract><doi>10.48550/arxiv.2102.05198</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2102.05198
ispartof
issn
language eng
recordid cdi_arxiv_primary_2102_05198
source arXiv.org
subjects Computer Science - Learning
Statistics - Machine Learning
title Statistical Inference for Polyak-Ruppert Averaged Zeroth-order Stochastic Gradient Algorithm
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-29T02%3A55%3A21IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Statistical%20Inference%20for%20Polyak-Ruppert%20Averaged%20Zeroth-order%20Stochastic%20Gradient%20Algorithm&rft.au=Jin,%20Yanhao&rft.date=2021-02-09&rft_id=info:doi/10.48550/arxiv.2102.05198&rft_dat=%3Carxiv_GOX%3E2102_05198%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true