{"id":"https://openalex.org/W3015508253","doi":"https://doi.org/10.1109/icassp40776.2020.9053864","title":"A Study of Generalization of Stochastic Mirror Descent Algorithms on Overparameterized Nonlinear Models","display_name":"A Study of Generalization of Stochastic Mirror Descent Algorithms on Overparameterized Nonlinear Models","publication_year":2020,"publication_date":"2020-04-09","ids":{"openalex":"https://openalex.org/W3015508253","doi":"https://doi.org/10.1109/icassp40776.2020.9053864","mag":"3015508253"},"language":"en","primary_location":{"id":"doi:10.1109/icassp40776.2020.9053864","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp40776.2020.9053864","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5005748450","display_name":"Navid Azizan","orcid":"https://orcid.org/0000-0002-4299-2963"},"institutions":[{"id":"https://openalex.org/I122411786","display_name":"California Institute of Technology","ror":"https://ror.org/05dxps055","country_code":"US","type":"education","lineage":["https://openalex.org/I122411786"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Navid Azizan","raw_affiliation_strings":["California Institute of Technology, Pasadena, CA"],"affiliations":[{"raw_affiliation_string":"California Institute of Technology, Pasadena, CA","institution_ids":["https://openalex.org/I122411786"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036507616","display_name":"Sahin Lale","orcid":"https://orcid.org/0000-0002-7191-346X"},"institutions":[{"id":"https://openalex.org/I122411786","display_name":"California Institute of Technology","ror":"https://ror.org/05dxps055","country_code":"US","type":"education","lineage":["https://openalex.org/I122411786"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sahin Lale","raw_affiliation_strings":["California Institute of Technology, Pasadena, CA"],"affiliations":[{"raw_affiliation_string":"California Institute of Technology, Pasadena, CA","institution_ids":["https://openalex.org/I122411786"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5002430773","display_name":"Babak Hassibi","orcid":"https://orcid.org/0000-0002-1375-5838"},"institutions":[{"id":"https://openalex.org/I122411786","display_name":"California Institute of Technology","ror":"https://ror.org/05dxps055","country_code":"US","type":"education","lineage":["https://openalex.org/I122411786"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Babak Hassibi","raw_affiliation_strings":["California Institute of Technology, Pasadena, CA"],"affiliations":[{"raw_affiliation_string":"California Institute of Technology, Pasadena, CA","institution_ids":["https://openalex.org/I122411786"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5005748450"],"corresponding_institution_ids":["https://openalex.org/I122411786"],"apc_list":null,"apc_paid":null,"fwci":0.6796,"has_fulltext":false,"cited_by_count":7,"citation_normalized_percentile":{"value":0.76029824,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":96},"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/T12676","display_name":"Machine Learning and ELM","score":0.9991999864578247,"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/T12676","display_name":"Machine Learning and ELM","score":0.9991999864578247,"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/T10057","display_name":"Face and Expression Recognition","score":0.9987000226974487,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9987000226974487,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/initialization","display_name":"Initialization","score":0.7778140306472778},{"id":"https://openalex.org/keywords/maxima-and-minima","display_name":"Maxima and minima","score":0.7206146717071533},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.6344982385635376},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.5551161170005798},{"id":"https://openalex.org/keywords/regularization","display_name":"Regularization (linguistics)","score":0.5546417236328125},{"id":"https://openalex.org/keywords/manifold","display_name":"Manifold (fluid mechanics)","score":0.538695752620697},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"Stochastic gradient descent","score":0.5377846360206604},{"id":"https://openalex.org/keywords/nonlinear-system","display_name":"Nonlinear system","score":0.5211408138275146},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.4629206657409668},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.4276401698589325},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.4209522008895874},{"id":"https://openalex.org/keywords/applied-mathematics","display_name":"Applied mathematics","score":0.38561007380485535},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.33240458369255066},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.324969083070755},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.22426342964172363},{"id":"https://openalex.org/keywords/mathematical-analysis","display_name":"Mathematical analysis","score":0.1380138397216797}],"concepts":[{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.7778140306472778},{"id":"https://openalex.org/C186633575","wikidata":"https://www.wikidata.org/wiki/Q845060","display_name":"Maxima and minima","level":2,"score":0.7206146717071533},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.6344982385635376},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.5551161170005798},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.5546417236328125},{"id":"https://openalex.org/C529865628","wikidata":"https://www.wikidata.org/wiki/Q1790740","display_name":"Manifold (fluid mechanics)","level":2,"score":0.538695752620697},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.5377846360206604},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.5211408138275146},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4629206657409668},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.4276401698589325},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4209522008895874},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.38561007380485535},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33240458369255066},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.324969083070755},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.22426342964172363},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.1380138397216797},{"id":"https://openalex.org/C50522688","wikidata":"https://www.wikidata.org/wiki/Q189833","display_name":"Economic growth","level":1,"score":0.0},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","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/C78519656","wikidata":"https://www.wikidata.org/wiki/Q101333","display_name":"Mechanical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","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/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/icassp40776.2020.9053864","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp40776.2020.9053864","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},{"id":"pmh:oai:authors.library.caltech.edu:102604","is_oa":false,"landing_page_url":"https://authors.library.caltech.edu/102604/","pdf_url":null,"source":{"id":"https://openalex.org/S4306402161","display_name":"CaltechAUTHORS (California Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I122411786","host_organization_name":"California Institute of Technology","host_organization_lineage":["https://openalex.org/I122411786"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"Book Section"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":33,"referenced_works":["https://openalex.org/W1505731132","https://openalex.org/W2016384870","https://openalex.org/W2095705004","https://openalex.org/W2118164131","https://openalex.org/W2194775991","https://openalex.org/W2613715972","https://openalex.org/W2768503813","https://openalex.org/W2776610675","https://openalex.org/W2806970110","https://openalex.org/W2899748887","https://openalex.org/W2905800572","https://openalex.org/W2911742574","https://openalex.org/W2938247478","https://openalex.org/W2948482439","https://openalex.org/W2962712496","https://openalex.org/W2962916883","https://openalex.org/W2963236897","https://openalex.org/W2963798163","https://openalex.org/W2964098911","https://openalex.org/W3118608800","https://openalex.org/W3137695714","https://openalex.org/W4299785371","https://openalex.org/W4301435544","https://openalex.org/W6674330103","https://openalex.org/W6737857923","https://openalex.org/W6745276634","https://openalex.org/W6746089872","https://openalex.org/W6746900977","https://openalex.org/W6748155593","https://openalex.org/W6748600614","https://openalex.org/W6752125765","https://openalex.org/W6763276191","https://openalex.org/W6763485134"],"related_works":["https://openalex.org/W2028661553","https://openalex.org/W2746420172","https://openalex.org/W2098344787","https://openalex.org/W2963959597","https://openalex.org/W2523060838","https://openalex.org/W2895642264","https://openalex.org/W4294351759","https://openalex.org/W2891160178","https://openalex.org/W4225513979","https://openalex.org/W3216123753"],"abstract_inverted_index":{"We":[0,122,147],"study":[1,164],"the":[2,4,8,21,27,36,57,71,85,90,96,104,107,131,135,141,166,178],"convergence,":[3],"implicit":[5,143],"regularization":[6,144],"and":[7],"generalization":[9,58,167],"of":[10,23,29,49,56,60,98,145,165,168],"stochastic":[11],"mirror":[12],"descent":[13],"(SMD)":[14],"algorithms":[15,73],"in":[16,84,106,137,157,213],"overparameterized":[17,86],"nonlinear":[18,87],"models,":[19],"where":[20,44],"number":[22,28],"model":[24],"parameters":[25],"exceeds":[26],"training":[30,37],"data":[31],"points.":[32],"Due":[33],"to":[34,66,95,118,134,204],"overpa-rameterization,":[35],"loss":[38],"has":[39],"infinitely":[40],"many":[41],"global":[42,69,99,120,127],"minima":[43,70],"they":[45],"define":[46],"a":[47,119,190],"manifold":[48,97],"interpolating":[50],"solutions.":[51],"To":[52],"have":[53],"an":[54,162,205],"understanding":[55],"performance":[59],"SMD":[61,72,116,169,181,217],"algorithms,":[62],"it":[63],"is":[64,92,102,129],"important":[65],"characterize":[67],"which":[68,101,212],"converge":[74],"to.":[75],"In":[76,171],"this":[77,126],"work,":[78],"we":[79,160,174],"first":[80],"theoretically":[81],"show":[82,175],"that":[83,125,151,176],"setting,":[88,110],"if":[89],"initialization":[91,136],"close":[93],"enough":[94],"minima,":[100],"usually":[103],"case":[105],"high":[108],"overparameterization":[109],"using":[111],"sufficiently":[112],"small":[113],"step":[114],"size,":[115],"converges":[117],"minimum.":[121],"further":[123],"prove":[124],"minimum":[128],"approximately":[130],"closest":[132],"one":[133],"Bregman":[138],"divergence,":[139],"demonstrating":[140],"approximate":[142],"SMD.":[146],"then":[148],"empirically":[149],"confirm":[150],"these":[152],"theoretical":[153],"results":[154],"are":[155],"observed":[156],"practice.":[158],"Finally,":[159],"provide":[161],"extensive":[163],"algorithms.":[170],"our":[172],"experiments,":[173],"on":[177],"CIFAR-10":[179],"dataset,":[180],"with":[182,218],"\u2113":[183,193,206,219],"<sub":[184,194,207,220],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[185,195,208,221],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">10</sub>":[186],"norm":[187,210,223],"potential":[188],"(as":[189],"surrogate":[191],"for":[192],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">\u221e</sub>":[196],")":[197],"consistently":[198,215],"general-izes":[199],"better":[200],"than":[201],"SGD":[202],"(corresponding":[203],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">2</sub>":[209],"potential),":[211],"turn":[214],"outperforms":[216],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">1</sub>":[222],"potential.":[224]},"counts_by_year":[{"year":2024,"cited_by_count":2},{"year":2023,"cited_by_count":2},{"year":2021,"cited_by_count":2},{"year":2020,"cited_by_count":1}],"updated_date":"2026-04-05T17:49:38.594831","created_date":"2025-10-10T00:00:00"}