{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,20]],"date-time":"2026-03-20T08:58:17Z","timestamp":1773997097474,"version":"3.50.1"},"reference-count":32,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,12,8]],"date-time":"2025-12-08T00:00:00Z","timestamp":1765152000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,12,8]],"date-time":"2025-12-08T00:00:00Z","timestamp":1765152000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,12,8]]},"DOI":"10.1109\/globecom59602.2025.11432271","type":"proceedings-article","created":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T20:04:01Z","timestamp":1773950641000},"page":"2663-2668","source":"Crossref","is-referenced-by-count":0,"title":["Asynchronous Decentralized Federated Learning Deconstructs Excessively Large Batches"],"prefix":"10.1109","author":[{"given":"Zixi","family":"Wang","sequence":"first","affiliation":[{"name":"Syracuse University,Department of Electrical Engineering and Computer Science,Syracuse,NY"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Feng","family":"Wang","sequence":"additional","affiliation":[{"name":"Syracuse University,Department of Electrical Engineering and Computer Science,Syracuse,NY"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"M. Cenk","family":"Gursoy","sequence":"additional","affiliation":[{"name":"Syracuse University,Department of Electrical Engineering and Computer Science,Syracuse,NY"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Senem","family":"Velipasalar","sequence":"additional","affiliation":[{"name":"Syracuse University,Department of Electrical Engineering and Computer Science,Syracuse,NY"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Federated optimization: Distributed machine learning for on-device intelligence","author":"Kone\u010dny","year":"2016"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/JIOT.2020.3030072"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1561\/2200000083"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/TMLCN.2024.3408131"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/JIOT.2024.3407584"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/COMST.2023.3315746"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1016\/j.array.2024.100360"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.23919\/ICN.2024.0008"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2020.2981035"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.4310\/amsa.2019.v4.n1.a1"},{"key":"ref11","article-title":"The anisotropic noise in stochastic gradient descent: Its behavior of escaping from sharp minima and regularization effects","author":"Zhu","year":"2018"},{"key":"ref12","article-title":"On the origin of implicit regularization in stochastic gradient descent","author":"Smith","year":"2021"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1118\/1.3213517"},{"key":"ref14","article-title":"Adding gradient noise improves learning for very deep networks","author":"Neelakantan","year":"2015"},{"key":"ref15","article-title":"Adam: A method for stochastic optimization","author":"Kingma","year":"2014"},{"key":"ref16","first-page":"1139","article-title":"On the importance of initialization and momentum in deep learning","volume-title":"International conference on machine learning","author":"Sutskever"},{"key":"ref17","article-title":"Fully decentralized federated learning","volume-title":"Third workshop on bayesian deep learning (NeurIPS)","volume":"2","author":"Lalitha"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2022.3174187"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2022.3198030"},{"key":"ref20","first-page":"3043","article-title":"Asynchronous decentralized parallel stochastic gradient descent","volume-title":"International Conference on Machine Learning","author":"Lian"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/TSIPN.2017.2695121"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/TCNS.2023.3242043"},{"key":"ref23","article-title":"On large-batch training for deep learning: Generalization gap and sharp minima","author":"Keskar","year":"2016"},{"key":"ref24","article-title":"Visualizing the loss landscape of neural nets","volume":"31","author":"Li","year":"2018","journal-title":"Advances in neural information processing systems"},{"key":"ref25","first-page":"2101","article-title":"Stochastic modified equations and adaptive stochastic gradient algorithms","volume-title":"International Conference on Machine Learning","author":"Li"},{"key":"ref26","first-page":"354","article-title":"A variational analysis of stochastic gradient algorithms","volume-title":"International conference on machine learning","author":"Mandt"},{"key":"ref27","article-title":"Three factors influencing minima in sgd","author":"Jastrz\u02dbebski","year":"2017"},{"key":"ref28","article-title":"Accurate, large minibatch sg d: training imagenet in 1 hour","author":"Goyal","year":"2017"},{"key":"ref29","article-title":"Large batch training of convolutional networks","author":"You","year":"2017"},{"key":"ref30","article-title":"Learning multiple layers of features from tiny images","author":"Krizhevsky","year":"2009"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00919"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"}],"event":{"name":"GLOBECOM 2025 - 2025 IEEE Global Communications Conference","location":"Taipei, Taiwan","start":{"date-parts":[[2025,12,8]]},"end":{"date-parts":[[2025,12,12]]}},"container-title":["GLOBECOM 2025 - 2025 IEEE Global Communications Conference"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11431620\/11431622\/11432271.pdf?arnumber=11432271","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,20]],"date-time":"2026-03-20T05:29:39Z","timestamp":1773984579000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11432271\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,12,8]]},"references-count":32,"URL":"https:\/\/doi.org\/10.1109\/globecom59602.2025.11432271","relation":{},"subject":[],"published":{"date-parts":[[2025,12,8]]}}}