When Computing Power Network Meets Distributed Machine Learning: An Efficient Federated Split Learning Framework
In this paper, we advocate CPN-FedSL, a novel and flexible Federated Split Learning (FedSL) framework over Computing Power Network (CPN). We build a dedicated model to capture the basic settings and learning characteristics (e.g., training flow, latency and convergence). Based on this model, we intr...
Gespeichert in:
Hauptverfasser: | , , , , , |
---|---|
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 | Yuan, Xinjing Pu, Lingjun Jiao, Lei Wang, Xiaofei Yang, Meijuan Xu, Jingdong |
description | In this paper, we advocate CPN-FedSL, a novel and flexible Federated Split
Learning (FedSL) framework over Computing Power Network (CPN). We build a
dedicated model to capture the basic settings and learning characteristics
(e.g., training flow, latency and convergence). Based on this model, we
introduce Resource Usage Effectiveness (RUE), a novel performance metric
integrating training utility with system cost, and formulate a multivariate
scheduling problem that maxi?mizes RUE by comprehensively taking client
admission, model partition, server selection, routing and bandwidth allocation
into account (i.e., mixed-integer fractional programming). We design Refinery,
an efficient approach that first linearizes the fractional objective and
non-convex constraints, and then solves the transformed problem via a greedy
based rounding algorithm in multiple iterations. Extensive evaluations
corroborate that CPN-FedSL is superior to the standard and state-of-the-art
learning frameworks (e.g., FedAvg and SplitFed), and besides Refinery is
lightweight and significantly outperforms its variants and de facto heuristic
methods under a variety of settings. |
doi_str_mv | 10.48550/arxiv.2305.12979 |
format | Article |
fullrecord | <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2305_12979</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2305_12979</sourcerecordid><originalsourceid>FETCH-LOGICAL-a679-e1860a02e46f4833eb5db26d13e3acc5e37707718851fb3e23bafcbc3369ec8c3</originalsourceid><addsrcrecordid>eNo9j7FOwzAURb0woMIHMOEfSLDjOHbYqtAAUgpIVGKMbOeZWjRO5LgU_r6kIKa7nHukg9AVJWkuOSc3Kny5zzRjhKc0K0V5jsa3LXhcDf24j86_45fhAAE_QTwM4QOvAeKE79wUg9P7CB1eK7N1HnADKvifwy1eeryy1hkHPuIaOghqBl_HnYv_GK6D6mF2XqAzq3YTXP7tAm3q1aZ6SJrn-8dq2SSqEGUCVBZEkQzywuaSMdC801nRUQZMGcOBCUGEoFJyajWDjGlljTaMFSUYadgCXf9qT8ntGFyvwnc7p7endHYEx4FVjA</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>When Computing Power Network Meets Distributed Machine Learning: An Efficient Federated Split Learning Framework</title><source>arXiv.org</source><creator>Yuan, Xinjing ; Pu, Lingjun ; Jiao, Lei ; Wang, Xiaofei ; Yang, Meijuan ; Xu, Jingdong</creator><creatorcontrib>Yuan, Xinjing ; Pu, Lingjun ; Jiao, Lei ; Wang, Xiaofei ; Yang, Meijuan ; Xu, Jingdong</creatorcontrib><description>In this paper, we advocate CPN-FedSL, a novel and flexible Federated Split
Learning (FedSL) framework over Computing Power Network (CPN). We build a
dedicated model to capture the basic settings and learning characteristics
(e.g., training flow, latency and convergence). Based on this model, we
introduce Resource Usage Effectiveness (RUE), a novel performance metric
integrating training utility with system cost, and formulate a multivariate
scheduling problem that maxi?mizes RUE by comprehensively taking client
admission, model partition, server selection, routing and bandwidth allocation
into account (i.e., mixed-integer fractional programming). We design Refinery,
an efficient approach that first linearizes the fractional objective and
non-convex constraints, and then solves the transformed problem via a greedy
based rounding algorithm in multiple iterations. Extensive evaluations
corroborate that CPN-FedSL is superior to the standard and state-of-the-art
learning frameworks (e.g., FedAvg and SplitFed), and besides Refinery is
lightweight and significantly outperforms its variants and de facto heuristic
methods under a variety of settings.</description><identifier>DOI: 10.48550/arxiv.2305.12979</identifier><language>eng</language><subject>Computer Science - Learning ; Computer Science - Networking and Internet Architecture</subject><creationdate>2023-05</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/2305.12979$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2305.12979$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Yuan, Xinjing</creatorcontrib><creatorcontrib>Pu, Lingjun</creatorcontrib><creatorcontrib>Jiao, Lei</creatorcontrib><creatorcontrib>Wang, Xiaofei</creatorcontrib><creatorcontrib>Yang, Meijuan</creatorcontrib><creatorcontrib>Xu, Jingdong</creatorcontrib><title>When Computing Power Network Meets Distributed Machine Learning: An Efficient Federated Split Learning Framework</title><description>In this paper, we advocate CPN-FedSL, a novel and flexible Federated Split
Learning (FedSL) framework over Computing Power Network (CPN). We build a
dedicated model to capture the basic settings and learning characteristics
(e.g., training flow, latency and convergence). Based on this model, we
introduce Resource Usage Effectiveness (RUE), a novel performance metric
integrating training utility with system cost, and formulate a multivariate
scheduling problem that maxi?mizes RUE by comprehensively taking client
admission, model partition, server selection, routing and bandwidth allocation
into account (i.e., mixed-integer fractional programming). We design Refinery,
an efficient approach that first linearizes the fractional objective and
non-convex constraints, and then solves the transformed problem via a greedy
based rounding algorithm in multiple iterations. Extensive evaluations
corroborate that CPN-FedSL is superior to the standard and state-of-the-art
learning frameworks (e.g., FedAvg and SplitFed), and besides Refinery is
lightweight and significantly outperforms its variants and de facto heuristic
methods under a variety of settings.</description><subject>Computer Science - Learning</subject><subject>Computer Science - Networking and Internet Architecture</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNo9j7FOwzAURb0woMIHMOEfSLDjOHbYqtAAUgpIVGKMbOeZWjRO5LgU_r6kIKa7nHukg9AVJWkuOSc3Kny5zzRjhKc0K0V5jsa3LXhcDf24j86_45fhAAE_QTwM4QOvAeKE79wUg9P7CB1eK7N1HnADKvifwy1eeryy1hkHPuIaOghqBl_HnYv_GK6D6mF2XqAzq3YTXP7tAm3q1aZ6SJrn-8dq2SSqEGUCVBZEkQzywuaSMdC801nRUQZMGcOBCUGEoFJyajWDjGlljTaMFSUYadgCXf9qT8ntGFyvwnc7p7endHYEx4FVjA</recordid><startdate>20230522</startdate><enddate>20230522</enddate><creator>Yuan, Xinjing</creator><creator>Pu, Lingjun</creator><creator>Jiao, Lei</creator><creator>Wang, Xiaofei</creator><creator>Yang, Meijuan</creator><creator>Xu, Jingdong</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20230522</creationdate><title>When Computing Power Network Meets Distributed Machine Learning: An Efficient Federated Split Learning Framework</title><author>Yuan, Xinjing ; Pu, Lingjun ; Jiao, Lei ; Wang, Xiaofei ; Yang, Meijuan ; Xu, Jingdong</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a679-e1860a02e46f4833eb5db26d13e3acc5e37707718851fb3e23bafcbc3369ec8c3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Computer Science - Learning</topic><topic>Computer Science - Networking and Internet Architecture</topic><toplevel>online_resources</toplevel><creatorcontrib>Yuan, Xinjing</creatorcontrib><creatorcontrib>Pu, Lingjun</creatorcontrib><creatorcontrib>Jiao, Lei</creatorcontrib><creatorcontrib>Wang, Xiaofei</creatorcontrib><creatorcontrib>Yang, Meijuan</creatorcontrib><creatorcontrib>Xu, Jingdong</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Yuan, Xinjing</au><au>Pu, Lingjun</au><au>Jiao, Lei</au><au>Wang, Xiaofei</au><au>Yang, Meijuan</au><au>Xu, Jingdong</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>When Computing Power Network Meets Distributed Machine Learning: An Efficient Federated Split Learning Framework</atitle><date>2023-05-22</date><risdate>2023</risdate><abstract>In this paper, we advocate CPN-FedSL, a novel and flexible Federated Split
Learning (FedSL) framework over Computing Power Network (CPN). We build a
dedicated model to capture the basic settings and learning characteristics
(e.g., training flow, latency and convergence). Based on this model, we
introduce Resource Usage Effectiveness (RUE), a novel performance metric
integrating training utility with system cost, and formulate a multivariate
scheduling problem that maxi?mizes RUE by comprehensively taking client
admission, model partition, server selection, routing and bandwidth allocation
into account (i.e., mixed-integer fractional programming). We design Refinery,
an efficient approach that first linearizes the fractional objective and
non-convex constraints, and then solves the transformed problem via a greedy
based rounding algorithm in multiple iterations. Extensive evaluations
corroborate that CPN-FedSL is superior to the standard and state-of-the-art
learning frameworks (e.g., FedAvg and SplitFed), and besides Refinery is
lightweight and significantly outperforms its variants and de facto heuristic
methods under a variety of settings.</abstract><doi>10.48550/arxiv.2305.12979</doi><oa>free_for_read</oa></addata></record> |
fulltext | fulltext_linktorsrc |
identifier | DOI: 10.48550/arxiv.2305.12979 |
ispartof | |
issn | |
language | eng |
recordid | cdi_arxiv_primary_2305_12979 |
source | arXiv.org |
subjects | Computer Science - Learning Computer Science - Networking and Internet Architecture |
title | When Computing Power Network Meets Distributed Machine Learning: An Efficient Federated Split Learning Framework |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T13%3A01%3A02IST&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=When%20Computing%20Power%20Network%20Meets%20Distributed%20Machine%20Learning:%20An%20Efficient%20Federated%20Split%20Learning%20Framework&rft.au=Yuan,%20Xinjing&rft.date=2023-05-22&rft_id=info:doi/10.48550/arxiv.2305.12979&rft_dat=%3Carxiv_GOX%3E2305_12979%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 |