{"id":"https://openalex.org/W3007957341","doi":null,"title":"Toward a theory of optimization for over-parameterized systems of non-linear equations: the lessons of deep learning.","display_name":"Toward a theory of optimization for over-parameterized systems of non-linear equations: the lessons of deep learning.","publication_year":2020,"publication_date":"2020-02-29","ids":{"openalex":"https://openalex.org/W3007957341","mag":"3007957341"},"language":"en","primary_location":{"id":"mag:3007957341","doi":null,"is_oa":false,"landing_page_url":"http://dblp.uni-trier.de/db/journals/corr/corr2003.html#abs-2003-00307","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":null},"volume":null,"type":"article","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Chaoyue Liu","orcid":"https://orcid.org/0000-0002-0301-1683","observed_orcids":["https://orcid.org/0000-0002-0301-1683"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chaoyue Liu","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0002-0301-1683","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090182859","display_name":"Libin Zhu","orcid":"https://orcid.org/0009-0009-0181-0105","observed_orcids":["https://orcid.org/0009-0009-0181-0105"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Libin Zhu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5014532691","display_name":"Mikhail A. Belkin","orcid":"https://orcid.org/0000-0003-3172-9462","observed_orcids":["https://orcid.org/0000-0003-3172-9462"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mikhail A. Belkin","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0003-3172-9462","affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":2.1703,"has_fulltext":false,"cited_by_count":33,"citation_normalized_percentile":{"value":0.89085109,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":99},"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9961000084877014,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9961000084877014,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.0013000000035390258,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10545","display_name":"Optimization and Variational Analysis","score":0.00039999998989515007,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/nonlinear-equations","display_name":"nonlinear equations","score":0.9240000247955322},{"id":"https://openalex.org/keywords/polyak-lojasiewicz-condition","display_name":"Polyak-Lojasiewicz condition","score":0.9419999718666077},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"gradient descent","score":0.6129999756813049},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"stochastic gradient descent","score":0.9750000238418579},{"id":"https://openalex.org/keywords/condition-number","display_name":"condition number","score":0.800000011920929},{"id":"https://openalex.org/keywords/nonconvex-optimization","display_name":"non-convex optimization","score":0.8379999995231628},{"id":"https://openalex.org/keywords/neural-network-training","display_name":"neural network training","score":0.8330000042915344},{"id":"https://openalex.org/keywords/convergence-analysis","display_name":"convergence analysis","score":0.578000009059906},{"id":"https://openalex.org/keywords/existence-of-solutions","display_name":"existence of solutions","score":0.871999979019165}],"concepts":[{"id":"https://openalex.org/C165464430","wikidata":"https://www.wikidata.org/wiki/Q1570441","display_name":"Parameterized complexity","level":2,"score":0.781101405620575},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.6627482771873474},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6500714421272278},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.561010479927063},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.5543439388275146},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.5453632473945618},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.528128445148468},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5276880264282227},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.5189499855041504},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.5088837146759033},{"id":"https://openalex.org/C138187205","wikidata":"https://www.wikidata.org/wiki/Q131251","display_name":"Tangent","level":2,"score":0.4842205047607422},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.45164769887924194},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4360804557800293},{"id":"https://openalex.org/C2777027219","wikidata":"https://www.wikidata.org/wiki/Q1284190","display_name":"Constant (computer programming)","level":2,"score":0.413411021232605},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35389864444732666},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2536800503730774},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.07660496234893799},{"id":"https://openalex.org/C118615104","wikidata":"https://www.wikidata.org/wiki/Q121416","display_name":"Discrete mathematics","level":1,"score":0.0664404034614563},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C50522688","wikidata":"https://www.wikidata.org/wiki/Q189833","display_name":"Economic growth","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:3007957341","doi":null,"is_oa":false,"landing_page_url":"http://dblp.uni-trier.de/db/journals/corr/corr2003.html#abs-2003-00307","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":31,"referenced_works":["https://openalex.org/W610261309","https://openalex.org/W2130076109","https://openalex.org/W2131172946","https://openalex.org/W2131703182","https://openalex.org/W2156779765","https://openalex.org/W2739160792","https://openalex.org/W2767286248","https://openalex.org/W2892675615","https://openalex.org/W2894604724","https://openalex.org/W2899748887","https://openalex.org/W2899874324","https://openalex.org/W2900959181","https://openalex.org/W2911867426","https://openalex.org/W2914270215","https://openalex.org/W2936995161","https://openalex.org/W2962698540","https://openalex.org/W2963248893","https://openalex.org/W2963518130","https://openalex.org/W2963834323","https://openalex.org/W2964121744","https://openalex.org/W2964467559","https://openalex.org/W2972392510","https://openalex.org/W2975935587","https://openalex.org/W2994295258","https://openalex.org/W3021189130","https://openalex.org/W3029645440","https://openalex.org/W3030163527","https://openalex.org/W3089465219","https://openalex.org/W3119866685","https://openalex.org/W3160443277","https://openalex.org/W4289334798"],"related_works":["https://openalex.org/W4226038297","https://openalex.org/W2899748887","https://openalex.org/W2963376662","https://openalex.org/W2782476368","https://openalex.org/W4415797779","https://openalex.org/W6986184950","https://openalex.org/W3118114660","https://openalex.org/W2894604724","https://openalex.org/W4405766326","https://openalex.org/W2979993567"],"abstract_inverted_index":{"The":[0],"success":[1],"of":[2,15,40,87,108,152,156],"deep":[3,47,122],"learning":[4],"is":[5],"due,":[6],"to":[7,11,20,58,90,95,105,118,133,149],"a":[8,43,101],"great":[9],"extent,":[10],"the":[12,69,84,91,96],"remarkable":[13],"effectiveness":[14],"gradient-based":[16],"optimization":[17,36,55,76],"methods":[18],"applied":[19],"large":[21,135],"neural":[22,48,123],"networks.":[23,49,124],"In":[24,50],"this":[25],"work":[26],"we":[27,52,126],"isolate":[28],"some":[29],"general":[30],"mathematical":[31],"structures":[32],"allowing":[33,73],"for":[34,74],"efficient":[35,75],"in":[37],"over-parameterized":[38,109],"systems":[39,60,89],"non-linear":[41,102],"equations,":[42],"setting":[44],"that":[45,54,129],"includes":[46],"particular,":[51],"show":[53],"problems":[56],"corresponding":[57],"such":[59],"are":[61],"not":[62],"convex,":[63],"even":[64,142],"locally,":[65],"but":[66],"instead":[67],"satisfy":[68],"Polyak-Lojasiewicz":[70],"(PL)":[71],"condition":[72,86,92],"by":[77],"gradient":[78,157],"descent":[79,158],"or":[80],"SGD.":[81,160],"We":[82,112],"connect":[83],"PL":[85],"these":[88,115],"number":[93],"associated":[94,132],"tangent":[97,130],"kernel":[98],"and":[99,121,154,159],"develop":[100],"theory":[103],"parallel":[104],"classical":[106],"analyses":[107],"linear":[110],"equations.":[111],"discuss":[113],"how":[114],"ideas":[116],"apply":[117],"training":[119],"shallow":[120],"Finally,":[125],"point":[127],"out":[128],"kernels":[131],"certain":[134],"system":[136],"may":[137],"be":[138],"far":[139],"from":[140],"constant,":[141],"locally.":[143],"Yet,":[144],"our":[145],"analysis":[146],"still":[147],"allows":[148],"demonstrate":[150],"existence":[151],"solutions":[153],"convergence":[155]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":6},{"year":2023,"cited_by_count":4},{"year":2022,"cited_by_count":3},{"year":2021,"cited_by_count":9},{"year":2020,"cited_by_count":10}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2020-03-06T00:00:00"}