PL-NMF: Parallel Locality-Optimized Non-negative Matrix Factorization

Non-negative Matrix Factorization (NMF) is a key kernel for unsupervised dimension reduction used in a wide range of applications, including topic modeling, recommender systems and bioinformatics. Due to the compute-intensive nature of applications that must perform repeated NMF, several parallel im...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Moon, Gordon E, Sukumaran-Rajam, Aravind, Parthasarathy, Srinivasan, Sadayappan, P
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 Moon, Gordon E
Sukumaran-Rajam, Aravind
Parthasarathy, Srinivasan
Sadayappan, P
description Non-negative Matrix Factorization (NMF) is a key kernel for unsupervised dimension reduction used in a wide range of applications, including topic modeling, recommender systems and bioinformatics. Due to the compute-intensive nature of applications that must perform repeated NMF, several parallel implementations have been developed in the past. However, existing parallel NMF algorithms have not addressed data locality optimizations, which are critical for high performance since data movement costs greatly exceed the cost of arithmetic/logic operations on current computer systems. In this paper, we devise a parallel NMF algorithm based on the HALS (Hierarchical Alternating Least Squares) scheme that incorporates algorithmic transformations to enhance data locality. Efficient realizations of the algorithm on multi-core CPUs and GPUs are developed, demonstrating significant performance improvement over existing state-of-the-art parallel NMF algorithms.
doi_str_mv 10.48550/arxiv.1904.07935
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_1904_07935</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1904_07935</sourcerecordid><originalsourceid>FETCH-LOGICAL-a675-5dac86037eaa2cf1e661529941e8535ef521811947b2022eaeca2c0c01fe6b5c3</originalsourceid><addsrcrecordid>eNotz8tqwzAUBFBtuihJP6Cr6gfk6uErW9mVELcF57HI3two10Wg2EEVIcnXN027GhiGgcPYs5JFWQPIV0zncCqUk2UhK2fgkS02rVgtmxnfYMIYKfJ29BhDvoj1MYdDuNKer8ZBDPSFOZyILzGncOYN-jymcL2V4zBlDz3Gb3r6zwnbNovt_EO06_fP-Vsr0FYgYI--ttJUhKh9r8haBdq5UlENBqgHrWqlXFnttNSakPxtJ71UPdkdeDNhL3-3d0d3TOGA6dL9erq7x_wAL2BFFg</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>PL-NMF: Parallel Locality-Optimized Non-negative Matrix Factorization</title><source>arXiv.org</source><creator>Moon, Gordon E ; Sukumaran-Rajam, Aravind ; Parthasarathy, Srinivasan ; Sadayappan, P</creator><creatorcontrib>Moon, Gordon E ; Sukumaran-Rajam, Aravind ; Parthasarathy, Srinivasan ; Sadayappan, P</creatorcontrib><description>Non-negative Matrix Factorization (NMF) is a key kernel for unsupervised dimension reduction used in a wide range of applications, including topic modeling, recommender systems and bioinformatics. Due to the compute-intensive nature of applications that must perform repeated NMF, several parallel implementations have been developed in the past. However, existing parallel NMF algorithms have not addressed data locality optimizations, which are critical for high performance since data movement costs greatly exceed the cost of arithmetic/logic operations on current computer systems. In this paper, we devise a parallel NMF algorithm based on the HALS (Hierarchical Alternating Least Squares) scheme that incorporates algorithmic transformations to enhance data locality. Efficient realizations of the algorithm on multi-core CPUs and GPUs are developed, demonstrating significant performance improvement over existing state-of-the-art parallel NMF algorithms.</description><identifier>DOI: 10.48550/arxiv.1904.07935</identifier><language>eng</language><subject>Computer Science - Distributed, Parallel, and Cluster Computing ; Computer Science - Learning ; Statistics - Machine Learning</subject><creationdate>2019-04</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>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/1904.07935$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.1904.07935$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Moon, Gordon E</creatorcontrib><creatorcontrib>Sukumaran-Rajam, Aravind</creatorcontrib><creatorcontrib>Parthasarathy, Srinivasan</creatorcontrib><creatorcontrib>Sadayappan, P</creatorcontrib><title>PL-NMF: Parallel Locality-Optimized Non-negative Matrix Factorization</title><description>Non-negative Matrix Factorization (NMF) is a key kernel for unsupervised dimension reduction used in a wide range of applications, including topic modeling, recommender systems and bioinformatics. Due to the compute-intensive nature of applications that must perform repeated NMF, several parallel implementations have been developed in the past. However, existing parallel NMF algorithms have not addressed data locality optimizations, which are critical for high performance since data movement costs greatly exceed the cost of arithmetic/logic operations on current computer systems. In this paper, we devise a parallel NMF algorithm based on the HALS (Hierarchical Alternating Least Squares) scheme that incorporates algorithmic transformations to enhance data locality. Efficient realizations of the algorithm on multi-core CPUs and GPUs are developed, demonstrating significant performance improvement over existing state-of-the-art parallel NMF algorithms.</description><subject>Computer Science - Distributed, Parallel, and Cluster Computing</subject><subject>Computer Science - Learning</subject><subject>Statistics - Machine Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2019</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz8tqwzAUBFBtuihJP6Cr6gfk6uErW9mVELcF57HI3two10Wg2EEVIcnXN027GhiGgcPYs5JFWQPIV0zncCqUk2UhK2fgkS02rVgtmxnfYMIYKfJ29BhDvoj1MYdDuNKer8ZBDPSFOZyILzGncOYN-jymcL2V4zBlDz3Gb3r6zwnbNovt_EO06_fP-Vsr0FYgYI--ttJUhKh9r8haBdq5UlENBqgHrWqlXFnttNSakPxtJ71UPdkdeDNhL3-3d0d3TOGA6dL9erq7x_wAL2BFFg</recordid><startdate>20190416</startdate><enddate>20190416</enddate><creator>Moon, Gordon E</creator><creator>Sukumaran-Rajam, Aravind</creator><creator>Parthasarathy, Srinivasan</creator><creator>Sadayappan, P</creator><scope>AKY</scope><scope>EPD</scope><scope>GOX</scope></search><sort><creationdate>20190416</creationdate><title>PL-NMF: Parallel Locality-Optimized Non-negative Matrix Factorization</title><author>Moon, Gordon E ; Sukumaran-Rajam, Aravind ; Parthasarathy, Srinivasan ; Sadayappan, P</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a675-5dac86037eaa2cf1e661529941e8535ef521811947b2022eaeca2c0c01fe6b5c3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2019</creationdate><topic>Computer Science - Distributed, Parallel, and Cluster Computing</topic><topic>Computer Science - Learning</topic><topic>Statistics - Machine Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Moon, Gordon E</creatorcontrib><creatorcontrib>Sukumaran-Rajam, Aravind</creatorcontrib><creatorcontrib>Parthasarathy, Srinivasan</creatorcontrib><creatorcontrib>Sadayappan, P</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>Moon, Gordon E</au><au>Sukumaran-Rajam, Aravind</au><au>Parthasarathy, Srinivasan</au><au>Sadayappan, P</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>PL-NMF: Parallel Locality-Optimized Non-negative Matrix Factorization</atitle><date>2019-04-16</date><risdate>2019</risdate><abstract>Non-negative Matrix Factorization (NMF) is a key kernel for unsupervised dimension reduction used in a wide range of applications, including topic modeling, recommender systems and bioinformatics. Due to the compute-intensive nature of applications that must perform repeated NMF, several parallel implementations have been developed in the past. However, existing parallel NMF algorithms have not addressed data locality optimizations, which are critical for high performance since data movement costs greatly exceed the cost of arithmetic/logic operations on current computer systems. In this paper, we devise a parallel NMF algorithm based on the HALS (Hierarchical Alternating Least Squares) scheme that incorporates algorithmic transformations to enhance data locality. Efficient realizations of the algorithm on multi-core CPUs and GPUs are developed, demonstrating significant performance improvement over existing state-of-the-art parallel NMF algorithms.</abstract><doi>10.48550/arxiv.1904.07935</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.1904.07935
ispartof
issn
language eng
recordid cdi_arxiv_primary_1904_07935
source arXiv.org
subjects Computer Science - Distributed, Parallel, and Cluster Computing
Computer Science - Learning
Statistics - Machine Learning
title PL-NMF: Parallel Locality-Optimized Non-negative Matrix Factorization
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-22T22%3A41%3A58IST&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=PL-NMF:%20Parallel%20Locality-Optimized%20Non-negative%20Matrix%20Factorization&rft.au=Moon,%20Gordon%20E&rft.date=2019-04-16&rft_id=info:doi/10.48550/arxiv.1904.07935&rft_dat=%3Carxiv_GOX%3E1904_07935%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