{"id":"https://openalex.org/W2965341779","doi":"https://doi.org/10.1137/19m129509x","title":"Extending the Step-Size Restriction for Gradient Descent to Avoid Strict Saddle Points","display_name":"Extending the Step-Size Restriction for Gradient Descent to Avoid Strict Saddle Points","publication_year":2020,"publication_date":"2020-01-01","ids":{"openalex":"https://openalex.org/W2965341779","doi":"https://doi.org/10.1137/19m129509x","mag":"2965341779"},"language":"en","primary_location":{"id":"doi:10.1137/19m129509x","is_oa":true,"landing_page_url":"https://doi.org/10.1137/19m129509x","pdf_url":null,"source":{"id":"https://openalex.org/S4210229561","display_name":"SIAM Journal on Mathematics of Data Science","issn_l":"2577-0187","issn":["2577-0187"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320508","host_organization_name":"Society for Industrial and Applied Mathematics","host_organization_lineage":["https://openalex.org/P4310320508"],"host_organization_lineage_names":["Society for Industrial and Applied Mathematics"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"SIAM Journal on Mathematics of Data Science","raw_type":"journal-article"},"type":"preprint","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.1137/19m129509x","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5086926394","display_name":"Hayden Schaeffer","orcid":"https://orcid.org/0000-0003-1379-1238"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Hayden Schaeffer","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5003184921","display_name":"Scott G. McCalla","orcid":"https://orcid.org/0000-0003-0093-4658"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Scott G. McCalla","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5086926394"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.1359,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.52229733,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":94},"biblio":{"volume":"2","issue":"4","first_page":"1181","last_page":"1197"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":1.0,"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":1.0,"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/T10500","display_name":"Sparse and Compressive Sensing Techniques","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/2206","display_name":"Computational Mechanics"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12072","display_name":"Machine Learning and Algorithms","score":0.9975000023841858,"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/saddle-point","display_name":"Saddle point","score":0.7984614372253418},{"id":"https://openalex.org/keywords/hessian-matrix","display_name":"Hessian matrix","score":0.7109811902046204},{"id":"https://openalex.org/keywords/lipschitz-continuity","display_name":"Lipschitz continuity","score":0.6988023519515991},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.6916114687919617},{"id":"https://openalex.org/keywords/constant","display_name":"Constant (computer programming)","score":0.6747109889984131},{"id":"https://openalex.org/keywords/initialization","display_name":"Initialization","score":0.6746349930763245},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"Gradient descent","score":0.6613346338272095},{"id":"https://openalex.org/keywords/saddle","display_name":"Saddle","score":0.5151016116142273},{"id":"https://openalex.org/keywords/differentiable-function","display_name":"Differentiable function","score":0.5023581981658936},{"id":"https://openalex.org/keywords/convex-function","display_name":"Convex function","score":0.49346768856048584},{"id":"https://openalex.org/keywords/limit","display_name":"Limit (mathematics)","score":0.48754116892814636},{"id":"https://openalex.org/keywords/regular-polygon","display_name":"Regular polygon","score":0.4554876387119293},{"id":"https://openalex.org/keywords/rate-of-convergence","display_name":"Rate of convergence","score":0.42919105291366577},{"id":"https://openalex.org/keywords/function","display_name":"Function (biology)","score":0.4148625135421753},{"id":"https://openalex.org/keywords/applied-mathematics","display_name":"Applied mathematics","score":0.39593613147735596},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.3378337025642395},{"id":"https://openalex.org/keywords/mathematical-analysis","display_name":"Mathematical analysis","score":0.31397250294685364},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.14353668689727783},{"id":"https://openalex.org/keywords/geometry","display_name":"Geometry","score":0.1271512806415558}],"concepts":[{"id":"https://openalex.org/C2681867","wikidata":"https://www.wikidata.org/wiki/Q690935","display_name":"Saddle point","level":2,"score":0.7984614372253418},{"id":"https://openalex.org/C203616005","wikidata":"https://www.wikidata.org/wiki/Q620495","display_name":"Hessian matrix","level":2,"score":0.7109811902046204},{"id":"https://openalex.org/C22324862","wikidata":"https://www.wikidata.org/wiki/Q652707","display_name":"Lipschitz continuity","level":2,"score":0.6988023519515991},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.6916114687919617},{"id":"https://openalex.org/C2777027219","wikidata":"https://www.wikidata.org/wiki/Q1284190","display_name":"Constant (computer programming)","level":2,"score":0.6747109889984131},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.6746349930763245},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.6613346338272095},{"id":"https://openalex.org/C2777127463","wikidata":"https://www.wikidata.org/wiki/Q10862618","display_name":"Saddle","level":2,"score":0.5151016116142273},{"id":"https://openalex.org/C202615002","wikidata":"https://www.wikidata.org/wiki/Q783507","display_name":"Differentiable function","level":2,"score":0.5023581981658936},{"id":"https://openalex.org/C145446738","wikidata":"https://www.wikidata.org/wiki/Q319913","display_name":"Convex function","level":3,"score":0.49346768856048584},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.48754116892814636},{"id":"https://openalex.org/C112680207","wikidata":"https://www.wikidata.org/wiki/Q714886","display_name":"Regular polygon","level":2,"score":0.4554876387119293},{"id":"https://openalex.org/C57869625","wikidata":"https://www.wikidata.org/wiki/Q1783502","display_name":"Rate of convergence","level":3,"score":0.42919105291366577},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.4148625135421753},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.39593613147735596},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3378337025642395},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.31397250294685364},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.14353668689727783},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.1271512806415558},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.0},{"id":"https://openalex.org/C78458016","wikidata":"https://www.wikidata.org/wiki/Q840400","display_name":"Evolutionary biology","level":1,"score":0.0},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.0},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"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/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.0}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1137/19m129509x","is_oa":true,"landing_page_url":"https://doi.org/10.1137/19m129509x","pdf_url":null,"source":{"id":"https://openalex.org/S4210229561","display_name":"SIAM Journal on Mathematics of Data Science","issn_l":"2577-0187","issn":["2577-0187"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320508","host_organization_name":"Society for Industrial and Applied Mathematics","host_organization_lineage":["https://openalex.org/P4310320508"],"host_organization_lineage_names":["Society for Industrial and Applied Mathematics"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"SIAM Journal on Mathematics of Data Science","raw_type":"journal-article"},{"id":"pmh:oai:arXiv.org:1908.01753","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1908.01753","pdf_url":"https://arxiv.org/pdf/1908.01753","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"mag:2965341779","is_oa":true,"landing_page_url":"https://arxiv.org/pdf/1908.01753.pdf","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},{"id":"doi:10.48550/arxiv.1908.01753","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.1908.01753","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.1137/19m129509x","is_oa":true,"landing_page_url":"https://doi.org/10.1137/19m129509x","pdf_url":null,"source":{"id":"https://openalex.org/S4210229561","display_name":"SIAM Journal on Mathematics of Data Science","issn_l":"2577-0187","issn":["2577-0187"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320508","host_organization_name":"Society for Industrial and Applied Mathematics","host_organization_lineage":["https://openalex.org/P4310320508"],"host_organization_lineage_names":["Society for Industrial and Applied Mathematics"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"SIAM Journal on Mathematics of Data Science","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1238848045","display_name":null,"funder_award_id":"FA9550-17-1-0125","funder_id":"https://openalex.org/F4320338279","funder_display_name":"Air Force Office of Scientific Research"},{"id":"https://openalex.org/G2801639669","display_name":null,"funder_award_id":"1752116","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G4374278986","display_name":null,"funder_award_id":"W911NF-19-1-0288","funder_id":"https://openalex.org/F4320338281","funder_display_name":"Army Research Office"},{"id":"https://openalex.org/G5761048035","display_name":null,"funder_award_id":"1813654","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320338279","display_name":"Air Force Office of Scientific Research","ror":"https://ror.org/011e9bt93"},{"id":"https://openalex.org/F4320338281","display_name":"Army Research Office","ror":"https://ror.org/05epdh915"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":26,"referenced_works":["https://openalex.org/W32532385","https://openalex.org/W1550918041","https://openalex.org/W1580352598","https://openalex.org/W1976222070","https://openalex.org/W2009941369","https://openalex.org/W2015836240","https://openalex.org/W2077612096","https://openalex.org/W2255205876","https://openalex.org/W2283023501","https://openalex.org/W2396019715","https://openalex.org/W2405479119","https://openalex.org/W2467441529","https://openalex.org/W2474090883","https://openalex.org/W2523060838","https://openalex.org/W2546420264","https://openalex.org/W2561193401","https://openalex.org/W2601846890","https://openalex.org/W2617691536","https://openalex.org/W2751090697","https://openalex.org/W2915116423","https://openalex.org/W2963075511","https://openalex.org/W2963417959","https://openalex.org/W2964106499","https://openalex.org/W2964265968","https://openalex.org/W3007399466","https://openalex.org/W3113425034"],"related_works":["https://openalex.org/W2971250964","https://openalex.org/W2912004605","https://openalex.org/W3025065608","https://openalex.org/W3006924802","https://openalex.org/W3198406029","https://openalex.org/W2889795689","https://openalex.org/W3175288054","https://openalex.org/W3147777701","https://openalex.org/W3002730424","https://openalex.org/W3086341836","https://openalex.org/W3133514242","https://openalex.org/W3083883621","https://openalex.org/W3158782800","https://openalex.org/W2253496675","https://openalex.org/W2133356919","https://openalex.org/W3022840531","https://openalex.org/W2799054451","https://openalex.org/W2791727399","https://openalex.org/W1614763381","https://openalex.org/W2751557393"],"abstract_inverted_index":{"We":[0,52,126],"provide":[1],"larger":[2],"step-size":[3,67],"restrictions":[4],"for":[5],"which":[6,138],"gradient":[7,26,63],"descent-based":[8],"algorithms":[9],"(almost":[10],"surely)":[11],"avoid":[12],"strict":[13,75],"saddle":[14,76],"points.":[15],"In":[16,100],"particular,":[17],"consider":[18],"a":[19,66,74,109,117,122],"twice":[20],"differentiable":[21],"(nonconvex)":[22],"objective":[23],"function":[24],"whose":[25],"has":[27],"Lipschitz":[28],"constant":[29,124],"$L$":[30],"and":[31],"that":[32,37,54,62,128,136],"the":[33,39,43,47,60,86,91,102,106,129,134,142],"set":[34],"of":[35,42,46],"points":[36],"obtain":[38],"maximum":[40],"value":[41],"spectral":[44],"norm":[45],"Hessian":[48],"is":[49,78,113],"measure":[50],"zero.":[51,79],"prove":[53],"given":[55],"one":[56],"uniformly":[57],"random":[58],"initialization,":[59],"probability":[61],"descent":[64],"with":[65,115,146],"up":[68,84],"to":[69,73,85,97,148],"$2/L$":[70],"will":[71],"converge":[72],"point":[77],"This":[80],"extends":[81],"previous":[82],"results":[83],"sharp":[87],"limit":[88],"imposed":[89],"by":[90],"convex":[92],"quadratic":[93],"case":[94,107],"(provably":[95],"converging":[96],"local":[98],"minimizers).":[99],"addition,":[101],"arguments":[103],"hold":[104],"in":[105,133],"when":[108],"learning":[110],"rate":[111,120],"schedule":[112],"given,":[114],"either":[116],"continuous":[118],"decaying":[119],"or":[121],"piecewise":[123],"schedule.":[125],"show":[127],"assumptions":[130,143],"are":[131,144],"robust":[132],"sense":[135],"functions":[137],"do":[139],"not":[140],"satisfy":[141],"meager":[145],"respect":[147],"analytic":[149],"functions.":[150]},"counts_by_year":[{"year":2023,"cited_by_count":1}],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2025-10-10T00:00:00"}