Neural Networks for Partially Linear Quantile Regression

Deep learning has enjoyed tremendous success in a variety of applications but its application to quantile regressions remains scarce. A major advantage of the deep learning approach is its flexibility to model complex data in a more parsimonious way than nonparametric smoothing methods. However, whi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhong, Qixian, Wang, Jane-Ling
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 Zhong, Qixian
Wang, Jane-Ling
description Deep learning has enjoyed tremendous success in a variety of applications but its application to quantile regressions remains scarce. A major advantage of the deep learning approach is its flexibility to model complex data in a more parsimonious way than nonparametric smoothing methods. However, while deep learning brought breakthroughs in prediction, it often lacks interpretability due to the black-box nature of multilayer structure with millions of parameters, hence it is not well suited for statistical inference. In this paper, we leverage the advantages of deep learning to apply it to quantile regression where the goal to produce interpretable results and perform statistical inference. We achieve this by adopting a semiparametric approach based on the partially linear quantile regression model, where covariates of primary interest for statistical inference are modelled linearly and all other covariates are modelled nonparametrically by means of a deep neural network. In addition to the new methodology, we provide theoretical justification for the proposed model by establishing the root-$n$ consistency and asymptotically normality of the parametric coefficient estimator and the minimax optimal convergence rate of the neural nonparametric function estimator. Across several simulated and real data examples, our proposed model empirically produces superior estimates and more accurate predictions than various alternative approaches.
doi_str_mv 10.48550/arxiv.2106.06225
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2106_06225</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2106_06225</sourcerecordid><originalsourceid>FETCH-arxiv_primary_2106_062253</originalsourceid><addsrcrecordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjQw0zMwMzIy5WSw8EstLUrMUfBLLSnPL8ouVkjLL1IISCwqyUzMyalU8MnMS00sUggsTcwrycxJVQhKTS9KLS7OzM_jYWBNS8wpTuWF0twM8m6uIc4eumA74guKMnMTiyrjQXbFg-0yJqwCACgfM_A</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Neural Networks for Partially Linear Quantile Regression</title><source>arXiv.org</source><creator>Zhong, Qixian ; Wang, Jane-Ling</creator><creatorcontrib>Zhong, Qixian ; Wang, Jane-Ling</creatorcontrib><description>Deep learning has enjoyed tremendous success in a variety of applications but its application to quantile regressions remains scarce. A major advantage of the deep learning approach is its flexibility to model complex data in a more parsimonious way than nonparametric smoothing methods. However, while deep learning brought breakthroughs in prediction, it often lacks interpretability due to the black-box nature of multilayer structure with millions of parameters, hence it is not well suited for statistical inference. In this paper, we leverage the advantages of deep learning to apply it to quantile regression where the goal to produce interpretable results and perform statistical inference. We achieve this by adopting a semiparametric approach based on the partially linear quantile regression model, where covariates of primary interest for statistical inference are modelled linearly and all other covariates are modelled nonparametrically by means of a deep neural network. In addition to the new methodology, we provide theoretical justification for the proposed model by establishing the root-$n$ consistency and asymptotically normality of the parametric coefficient estimator and the minimax optimal convergence rate of the neural nonparametric function estimator. Across several simulated and real data examples, our proposed model empirically produces superior estimates and more accurate predictions than various alternative approaches.</description><identifier>DOI: 10.48550/arxiv.2106.06225</identifier><language>eng</language><subject>Mathematics - Statistics Theory ; Statistics - Theory</subject><creationdate>2021-06</creationdate><rights>http://creativecommons.org/licenses/by/4.0</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/2106.06225$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2106.06225$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Zhong, Qixian</creatorcontrib><creatorcontrib>Wang, Jane-Ling</creatorcontrib><title>Neural Networks for Partially Linear Quantile Regression</title><description>Deep learning has enjoyed tremendous success in a variety of applications but its application to quantile regressions remains scarce. A major advantage of the deep learning approach is its flexibility to model complex data in a more parsimonious way than nonparametric smoothing methods. However, while deep learning brought breakthroughs in prediction, it often lacks interpretability due to the black-box nature of multilayer structure with millions of parameters, hence it is not well suited for statistical inference. In this paper, we leverage the advantages of deep learning to apply it to quantile regression where the goal to produce interpretable results and perform statistical inference. We achieve this by adopting a semiparametric approach based on the partially linear quantile regression model, where covariates of primary interest for statistical inference are modelled linearly and all other covariates are modelled nonparametrically by means of a deep neural network. In addition to the new methodology, we provide theoretical justification for the proposed model by establishing the root-$n$ consistency and asymptotically normality of the parametric coefficient estimator and the minimax optimal convergence rate of the neural nonparametric function estimator. Across several simulated and real data examples, our proposed model empirically produces superior estimates and more accurate predictions than various alternative approaches.</description><subject>Mathematics - Statistics Theory</subject><subject>Statistics - Theory</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjQw0zMwMzIy5WSw8EstLUrMUfBLLSnPL8ouVkjLL1IISCwqyUzMyalU8MnMS00sUggsTcwrycxJVQhKTS9KLS7OzM_jYWBNS8wpTuWF0twM8m6uIc4eumA74guKMnMTiyrjQXbFg-0yJqwCACgfM_A</recordid><startdate>20210611</startdate><enddate>20210611</enddate><creator>Zhong, Qixian</creator><creator>Wang, Jane-Ling</creator><scope>AKZ</scope><scope>EPD</scope><scope>GOX</scope></search><sort><creationdate>20210611</creationdate><title>Neural Networks for Partially Linear Quantile Regression</title><author>Zhong, Qixian ; Wang, Jane-Ling</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-arxiv_primary_2106_062253</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Mathematics - Statistics Theory</topic><topic>Statistics - Theory</topic><toplevel>online_resources</toplevel><creatorcontrib>Zhong, Qixian</creatorcontrib><creatorcontrib>Wang, Jane-Ling</creatorcontrib><collection>arXiv Mathematics</collection><collection>arXiv Statistics</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Zhong, Qixian</au><au>Wang, Jane-Ling</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Neural Networks for Partially Linear Quantile Regression</atitle><date>2021-06-11</date><risdate>2021</risdate><abstract>Deep learning has enjoyed tremendous success in a variety of applications but its application to quantile regressions remains scarce. A major advantage of the deep learning approach is its flexibility to model complex data in a more parsimonious way than nonparametric smoothing methods. However, while deep learning brought breakthroughs in prediction, it often lacks interpretability due to the black-box nature of multilayer structure with millions of parameters, hence it is not well suited for statistical inference. In this paper, we leverage the advantages of deep learning to apply it to quantile regression where the goal to produce interpretable results and perform statistical inference. We achieve this by adopting a semiparametric approach based on the partially linear quantile regression model, where covariates of primary interest for statistical inference are modelled linearly and all other covariates are modelled nonparametrically by means of a deep neural network. In addition to the new methodology, we provide theoretical justification for the proposed model by establishing the root-$n$ consistency and asymptotically normality of the parametric coefficient estimator and the minimax optimal convergence rate of the neural nonparametric function estimator. Across several simulated and real data examples, our proposed model empirically produces superior estimates and more accurate predictions than various alternative approaches.</abstract><doi>10.48550/arxiv.2106.06225</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2106.06225
ispartof
issn
language eng
recordid cdi_arxiv_primary_2106_06225
source arXiv.org
subjects Mathematics - Statistics Theory
Statistics - Theory
title Neural Networks for Partially Linear Quantile Regression
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-19T10%3A50%3A57IST&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=Neural%20Networks%20for%20Partially%20Linear%20Quantile%20Regression&rft.au=Zhong,%20Qixian&rft.date=2021-06-11&rft_id=info:doi/10.48550/arxiv.2106.06225&rft_dat=%3Carxiv_GOX%3E2106_06225%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