pISTA: preconditioned Iterative Soft Thresholding Algorithm for Graphical Lasso

We propose a novel quasi-Newton method for solving the sparse inverse covariance estimation problem also known as the graphical least absolute shrinkage and selection operator (GLASSO). This problem is often solved using a second-order quadratic approximation. However, in such algorithms the Hessian...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Shalom, Gal, Treister, Eran, Yavneh, Irad
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 Shalom, Gal
Treister, Eran
Yavneh, Irad
description We propose a novel quasi-Newton method for solving the sparse inverse covariance estimation problem also known as the graphical least absolute shrinkage and selection operator (GLASSO). This problem is often solved using a second-order quadratic approximation. However, in such algorithms the Hessian term is complex and computationally expensive to handle. Therefore, our method uses the inverse of the Hessian as a preconditioner to simplify and approximate the quadratic element at the cost of a more complex \(\ell_1\) element. The variables of the resulting preconditioned problem are coupled only by the \(\ell_1\) sub-derivative of each other, which can be guessed with minimal cost using the gradient itself, allowing the algorithm to be parallelized and implemented efficiently on GPU hardware accelerators. Numerical results on synthetic and real data demonstrate that our method is competitive with other state-of-the-art approaches.
doi_str_mv 10.48550/arxiv.2205.10027
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2205_10027</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2205_10027</sourcerecordid><originalsourceid>FETCH-LOGICAL-a677-4b117374c80e3d149a1a7414a8c207b13f02df4d2fe448e1b34b3df4b8946a9e3</originalsourceid><addsrcrecordid>eNotz71qwzAYhWEtGUrSC-hU3YBd_XyO7G4mtKnBkCHezSdLigWOZWQR2rtvm3Y68A4HHkKeOMuhLAr2gvHT33IhWJFzxoR6IKelOXf1K12iHcJsfPJhtoY2yUZM_mbpObhEuzHadQyT8fOF1tMlRJ_GK3Uh0mPEZfQDTrTFdQ07snE4rfbxf7eke3_rDh9Zezo2h7rNcK9UBppzJRUMJbPScKiQowIOWA6CKc2lY8I4MMJZgNJyLUHLn6DLCvZYWbklz3-3d1G_RH_F-NX_yvq7TH4DwSFI4w</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>pISTA: preconditioned Iterative Soft Thresholding Algorithm for Graphical Lasso</title><source>arXiv.org</source><creator>Shalom, Gal ; Treister, Eran ; Yavneh, Irad</creator><creatorcontrib>Shalom, Gal ; Treister, Eran ; Yavneh, Irad</creatorcontrib><description>We propose a novel quasi-Newton method for solving the sparse inverse covariance estimation problem also known as the graphical least absolute shrinkage and selection operator (GLASSO). This problem is often solved using a second-order quadratic approximation. However, in such algorithms the Hessian term is complex and computationally expensive to handle. Therefore, our method uses the inverse of the Hessian as a preconditioner to simplify and approximate the quadratic element at the cost of a more complex \(\ell_1\) element. The variables of the resulting preconditioned problem are coupled only by the \(\ell_1\) sub-derivative of each other, which can be guessed with minimal cost using the gradient itself, allowing the algorithm to be parallelized and implemented efficiently on GPU hardware accelerators. Numerical results on synthetic and real data demonstrate that our method is competitive with other state-of-the-art approaches.</description><identifier>DOI: 10.48550/arxiv.2205.10027</identifier><language>eng</language><subject>Computer Science - Mathematical Software ; Computer Science - Numerical Analysis ; Mathematics - Numerical Analysis</subject><creationdate>2022-05</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>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2205.10027$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2205.10027$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Shalom, Gal</creatorcontrib><creatorcontrib>Treister, Eran</creatorcontrib><creatorcontrib>Yavneh, Irad</creatorcontrib><title>pISTA: preconditioned Iterative Soft Thresholding Algorithm for Graphical Lasso</title><description>We propose a novel quasi-Newton method for solving the sparse inverse covariance estimation problem also known as the graphical least absolute shrinkage and selection operator (GLASSO). This problem is often solved using a second-order quadratic approximation. However, in such algorithms the Hessian term is complex and computationally expensive to handle. Therefore, our method uses the inverse of the Hessian as a preconditioner to simplify and approximate the quadratic element at the cost of a more complex \(\ell_1\) element. The variables of the resulting preconditioned problem are coupled only by the \(\ell_1\) sub-derivative of each other, which can be guessed with minimal cost using the gradient itself, allowing the algorithm to be parallelized and implemented efficiently on GPU hardware accelerators. Numerical results on synthetic and real data demonstrate that our method is competitive with other state-of-the-art approaches.</description><subject>Computer Science - Mathematical Software</subject><subject>Computer Science - Numerical Analysis</subject><subject>Mathematics - Numerical Analysis</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz71qwzAYhWEtGUrSC-hU3YBd_XyO7G4mtKnBkCHezSdLigWOZWQR2rtvm3Y68A4HHkKeOMuhLAr2gvHT33IhWJFzxoR6IKelOXf1K12iHcJsfPJhtoY2yUZM_mbpObhEuzHadQyT8fOF1tMlRJ_GK3Uh0mPEZfQDTrTFdQ07snE4rfbxf7eke3_rDh9Zezo2h7rNcK9UBppzJRUMJbPScKiQowIOWA6CKc2lY8I4MMJZgNJyLUHLn6DLCvZYWbklz3-3d1G_RH_F-NX_yvq7TH4DwSFI4w</recordid><startdate>20220520</startdate><enddate>20220520</enddate><creator>Shalom, Gal</creator><creator>Treister, Eran</creator><creator>Yavneh, Irad</creator><scope>AKY</scope><scope>AKZ</scope><scope>GOX</scope></search><sort><creationdate>20220520</creationdate><title>pISTA: preconditioned Iterative Soft Thresholding Algorithm for Graphical Lasso</title><author>Shalom, Gal ; Treister, Eran ; Yavneh, Irad</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a677-4b117374c80e3d149a1a7414a8c207b13f02df4d2fe448e1b34b3df4b8946a9e3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Computer Science - Mathematical Software</topic><topic>Computer Science - Numerical Analysis</topic><topic>Mathematics - Numerical Analysis</topic><toplevel>online_resources</toplevel><creatorcontrib>Shalom, Gal</creatorcontrib><creatorcontrib>Treister, Eran</creatorcontrib><creatorcontrib>Yavneh, Irad</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv Mathematics</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Shalom, Gal</au><au>Treister, Eran</au><au>Yavneh, Irad</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>pISTA: preconditioned Iterative Soft Thresholding Algorithm for Graphical Lasso</atitle><date>2022-05-20</date><risdate>2022</risdate><abstract>We propose a novel quasi-Newton method for solving the sparse inverse covariance estimation problem also known as the graphical least absolute shrinkage and selection operator (GLASSO). This problem is often solved using a second-order quadratic approximation. However, in such algorithms the Hessian term is complex and computationally expensive to handle. Therefore, our method uses the inverse of the Hessian as a preconditioner to simplify and approximate the quadratic element at the cost of a more complex \(\ell_1\) element. The variables of the resulting preconditioned problem are coupled only by the \(\ell_1\) sub-derivative of each other, which can be guessed with minimal cost using the gradient itself, allowing the algorithm to be parallelized and implemented efficiently on GPU hardware accelerators. Numerical results on synthetic and real data demonstrate that our method is competitive with other state-of-the-art approaches.</abstract><doi>10.48550/arxiv.2205.10027</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2205.10027
ispartof
issn
language eng
recordid cdi_arxiv_primary_2205_10027
source arXiv.org
subjects Computer Science - Mathematical Software
Computer Science - Numerical Analysis
Mathematics - Numerical Analysis
title pISTA: preconditioned Iterative Soft Thresholding Algorithm for Graphical Lasso
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-06T16%3A28%3A16IST&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=pISTA:%20preconditioned%20Iterative%20Soft%20Thresholding%20Algorithm%20for%20Graphical%20Lasso&rft.au=Shalom,%20Gal&rft.date=2022-05-20&rft_id=info:doi/10.48550/arxiv.2205.10027&rft_dat=%3Carxiv_GOX%3E2205_10027%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