{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,23]],"date-time":"2026-07-23T04:00:16Z","timestamp":1784779216570,"version":"3.55.0"},"reference-count":61,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"6","license":[{"start":{"date-parts":[[2019,6,1]],"date-time":"2019-06-01T00:00:00Z","timestamp":1559347200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2019,6,1]],"date-time":"2019-06-01T00:00:00Z","timestamp":1559347200000},"content-version":"am","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2019,6,1]],"date-time":"2019-06-01T00:00:00Z","timestamp":1559347200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2019,6,1]],"date-time":"2019-06-01T00:00:00Z","timestamp":1559347200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"DARPA Young Faculty Award","award":["N66001-14-1-4047"],"award-info":[{"award-number":["N66001-14-1-4047"]}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["DMS-1454377-CAREER"],"award-info":[{"award-number":["DMS-1454377-CAREER"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["DMS-1841569-CAREER"],"award-info":[{"award-number":["DMS-1841569-CAREER"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["BIGDATA-1546482"],"award-info":[{"award-number":["BIGDATA-1546482"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["TRIPODS-1740735"],"award-info":[{"award-number":["TRIPODS-1740735"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["BIGDATA-1840866"],"award-info":[{"award-number":["BIGDATA-1840866"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["RI-1408910"],"award-info":[{"award-number":["RI-1408910"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["IIS-1332109"],"award-info":[{"award-number":["IIS-1332109"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"name":"NIH","award":["R01MH102339"],"award-info":[{"award-number":["R01MH102339"]}]},{"name":"NIH","award":["R01GM083084"],"award-info":[{"award-number":["R01GM083084"]}]},{"name":"Alfred P. Sloan Fellowship"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Inform. Theory"],"published-print":{"date-parts":[[2019,6]]},"DOI":"10.1109\/tit.2019.2898663","type":"journal-article","created":{"date-parts":[[2019,3,27]],"date-time":"2019-03-27T22:02:42Z","timestamp":1553724162000},"page":"3489-3514","source":"Crossref","is-referenced-by-count":20,"title":["Symmetry, Saddle Points, and Global Optimization Landscape of Nonconvex Matrix Factorization"],"prefix":"10.1109","volume":"65","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5510-9447","authenticated-orcid":false,"given":"Xingguo","family":"Li","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Junwei","family":"Lu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Raman","family":"Arora","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1570-7071","authenticated-orcid":false,"given":"Jarvis","family":"Haupt","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Han","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhaoran","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4991-7851","authenticated-orcid":false,"given":"Tuo","family":"Zhao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2015.2401536"},{"key":"ref38","author":"haupt","year":"2017","journal-title":"Near optimal sketching of low-rank tensor regression"},{"key":"ref33","author":"lin","year":"2016","journal-title":"A non-convex one-pass framework for generalized factorization machine and rank-one matrix sensing"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1214\/13-AOS1151"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1214\/12-AOS1014"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1214\/13-AOS1178"},{"key":"ref37","first-page":"2238","article-title":"Contrastive learning using spectral methods","author":"zou","year":"2013","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref36","first-page":"2375","article-title":"Provable ICA with unknown Gaussian noise, with implications for Gaussian mixtures and autoencoders","author":"arora","year":"2012","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1214\/14-AOS1208"},{"key":"ref34","author":"blondel","year":"2016","journal-title":"Polynomial networks and factorization machines New insights and efficient training algorithms"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1007\/978-0-8176-4948-7"},{"key":"ref61","author":"vershynin","year":"2010","journal-title":"Introduction to the Non-Asymptotic Analysis of Random Matrices"},{"key":"ref28","author":"ge","year":"2016","journal-title":"Matrix completion has no spurious local minimum"},{"key":"ref27","author":"zheng","year":"2016","journal-title":"Convergence analysis for rectangular matrix completion using burer-monteiro factorization and gradient descent"},{"key":"ref29","author":"jin","year":"2016","journal-title":"Provable efficient online matrix completion via non-convex stochastic gradient descent"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1137\/070697835"},{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1007\/s10208-009-9045-5"},{"key":"ref20","first-page":"559","article-title":"A nonconvex optimization framework for low rank matrix estimation","author":"zhao","year":"2015","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref22","author":"bhojanapalli","year":"2016","journal-title":"Global optimality of local search for low rank matrix recovery"},{"key":"ref21","author":"tu","year":"2015","journal-title":"Low-rank solutions of linear matrix equations via procrustes flow"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/ISIT.2009.5205567"},{"key":"ref23","author":"park","year":"2016","journal-title":"Non-square matrix sensing without spurious local minima via the Burer-Monteiro approach"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/FOCS.2015.25"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/FOCS.2014.75"},{"key":"ref50","volume":"3","author":"dummit","year":"2004","journal-title":"Abstract Algebra"},{"key":"ref51","author":"goodfellow","year":"2016","journal-title":"Deep Learning"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2011.2111771"},{"key":"ref58","first-page":"1665","article-title":"Restricted strong convexity and weighted matrix completion: Optimal bounds with noise","volume":"13","author":"negahban","year":"2012","journal-title":"J Mach Learn Res"},{"key":"ref57","first-page":"777","article-title":"Online partial least square optimization: Dropping convexity for better efficiency and scalability","volume":"70","author":"chen","year":"2017","journal-title":"Proc 34th Int Conf Mach Learn"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46128-1_50"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1016\/0041-5553(63)90382-3"},{"key":"ref54","author":"carmon","year":"2016","journal-title":"Gradient descent efficiently finds the cubic-regularized non-convex newton step"},{"key":"ref53","first-page":"720","article-title":"Weighted low-rank approximations","author":"srebro","year":"2003","journal-title":"Proc 20th Int Conf Mach Learn"},{"key":"ref52","article-title":"Introduction to differential geometry","author":"robbin","year":"2011","journal-title":"ETH Lect Notes Preliminary Version"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1007\/s10107-004-0564-1"},{"key":"ref11","first-page":"530","article-title":"Dropping convexity for faster semi-definite optimization","author":"bhojanapalli","year":"2016","journal-title":"Proc Conf Learn Theory"},{"key":"ref40","author":"yi","year":"2016","journal-title":"Fast algorithms for robust PCA via gradient descent"},{"key":"ref12","first-page":"797","article-title":"Escaping from saddle points—Online stochastic gradient for tensor decomposition","author":"ge","year":"2015","journal-title":"Proc 28th Conf Learn Theory"},{"key":"ref13","author":"chen","year":"2015","journal-title":"Fast low-rank estimation by projected gradient descent General statistical and algorithmic guarantees"},{"key":"ref14","author":"anandkumar","year":"2016","journal-title":"Efficient approaches for escaping higher order saddle points in non-convex optimization"},{"key":"ref15","author":"park","year":"2016","journal-title":"Finding low-rank solutions via non-convex matrix factorization efficiently and provably"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/3055399.3055464"},{"key":"ref17","author":"allen-zhu","year":"0","journal-title":"Natasha 2 Faster non-convex optimization than SGD"},{"key":"ref18","author":"liu","year":"2017","journal-title":"On noisy negative curvature descent Competing with gradient descent for faster non-convex optimization"},{"key":"ref19","author":"yang","year":"2016","journal-title":"Unified convergence analysis of stochastic momentum methods for convex and non-convex optimization"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1007\/s12532-011-0029-5"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1137\/080738970"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2010.2054251"},{"key":"ref5","first-page":"937","article-title":"Guaranteed rank minimization via singular value projection","author":"jain","year":"2010","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref8","author":"li","year":"2016","journal-title":"Nonconvex sparse learning via stochastic optimization with progressive variance reduction"},{"key":"ref7","author":"shalev-shwartz","year":"2011","journal-title":"Large-scale convex minimization with a low-rank constraint"},{"key":"ref49","author":"ge","year":"2017","journal-title":"No spurious local minima in nonconvex low rank problems A unified geometric analysis"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1007\/s10107-002-0352-8"},{"key":"ref46","author":"panageas","year":"2016","journal-title":"Gradient Descent Only Converges to Minimizers Non-Isolated Critical Points and Invariant Regions"},{"key":"ref45","first-page":"1246","article-title":"Gradient descent only converges to minimizers","volume":"49","author":"lee","year":"2016","journal-title":"Proc 29th Annu Conf Learn Theory (Proceedings of Machine Learning Research)"},{"key":"ref48","author":"chen","year":"2018","journal-title":"On landscape of Lagrangian functions and stochastic search for constrained nonconvex optimization"},{"key":"ref47","author":"zhu","year":"2017","journal-title":"The global optimization geometry of low-rank matrix optimization"},{"key":"ref42","first-page":"109","article-title":"A convergent gradient descent algorithm for rank minimization and semidefinite programming from random linear measurements","author":"zheng","year":"2015","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref41","first-page":"600","article-title":"Low-rank and sparse structure pursuit via alternating minimization","author":"gu","year":"2016","journal-title":"Proc 19th Int Conf Artif Intell"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1109\/ISIT.2016.7541725"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2015.2399924"}],"container-title":["IEEE Transactions on Information Theory"],"original-title":[],"link":[{"URL":"https:\/\/ieeexplore.ieee.org\/ielaam\/18\/8718425\/8675509-aam.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/18\/8718425\/08675509.pdf?arnumber=8675509","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,7,13]],"date-time":"2022-07-13T20:43:10Z","timestamp":1657744990000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/8675509\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2019,6]]},"references-count":61,"journal-issue":{"issue":"6"},"URL":"https:\/\/doi.org\/10.1109\/tit.2019.2898663","relation":{},"ISSN":["0018-9448","1557-9654"],"issn-type":[{"value":"0018-9448","type":"print"},{"value":"1557-9654","type":"electronic"}],"subject":[],"published":{"date-parts":[[2019,6]]}}}