{"id":"https://openalex.org/W4407367788","doi":"https://doi.org/10.1007/s10589-025-00656-1","title":"On the convergence of the gradient descent method with stochastic fixed-point rounding errors under the Polyak\u2013\u0141ojasiewicz inequality","display_name":"On the convergence of the gradient descent method with stochastic fixed-point rounding errors under the Polyak\u2013\u0141ojasiewicz inequality","publication_year":2025,"publication_date":"2025-02-11","ids":{"openalex":"https://openalex.org/W4407367788","doi":"https://doi.org/10.1007/s10589-025-00656-1"},"language":"en","primary_location":{"id":"doi:10.1007/s10589-025-00656-1","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s10589-025-00656-1","pdf_url":"https://link.springer.com/content/pdf/10.1007/s10589-025-00656-1.pdf","source":{"id":"https://openalex.org/S39243548","display_name":"Computational Optimization and Applications","issn_l":"0926-6003","issn":["0926-6003","1573-2894"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Computational Optimization and Applications","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://link.springer.com/content/pdf/10.1007/s10589-025-00656-1.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5064471935","display_name":"Lu Xia","orcid":null},"institutions":[{"id":"https://openalex.org/I83019370","display_name":"Eindhoven University of Technology","ror":"https://ror.org/02c2kyt77","country_code":"NL","type":"education","lineage":["https://openalex.org/I83019370"]}],"countries":["NL"],"is_corresponding":true,"raw_author_name":"Lu Xia","raw_affiliation_strings":["Department of Mathematics and Computer Science, Eindhoven University of Technology, 5600 MB, Eindhoven, The Netherlands"],"raw_orcid":"https://orcid.org/0000-0001-5784-0367","affiliations":[{"raw_affiliation_string":"Department of Mathematics and Computer Science, Eindhoven University of Technology, 5600 MB, Eindhoven, The Netherlands","institution_ids":["https://openalex.org/I83019370"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030708699","display_name":"Stefano Massei","orcid":"https://orcid.org/0000-0003-1813-4181"},"institutions":[{"id":"https://openalex.org/I108290504","display_name":"University of Pisa","ror":"https://ror.org/03ad39j10","country_code":"IT","type":"education","lineage":["https://openalex.org/I108290504"]}],"countries":["IT"],"is_corresponding":false,"raw_author_name":"Stefano Massei","raw_affiliation_strings":["Department of Mathematics, Universit\u00e0 di Pisa, Pisa, 56127, Italy"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Mathematics, Universit\u00e0 di Pisa, Pisa, 56127, Italy","institution_ids":["https://openalex.org/I108290504"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5064293224","display_name":"Michiel E. Hochstenbach","orcid":"https://orcid.org/0000-0002-9196-8257"},"institutions":[{"id":"https://openalex.org/I83019370","display_name":"Eindhoven University of Technology","ror":"https://ror.org/02c2kyt77","country_code":"NL","type":"education","lineage":["https://openalex.org/I83019370"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Michiel E. Hochstenbach","raw_affiliation_strings":["Department of Mathematics and Computer Science, Eindhoven University of Technology, 5600 MB, Eindhoven, The Netherlands"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Mathematics and Computer Science, Eindhoven University of Technology, 5600 MB, Eindhoven, The Netherlands","institution_ids":["https://openalex.org/I83019370"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5064471935"],"corresponding_institution_ids":["https://openalex.org/I83019370"],"apc_list":{"value":2390,"currency":"EUR","value_usd":2990},"apc_paid":{"value":2390,"currency":"EUR","value_usd":2990},"fwci":2.792,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.89743933,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":95,"max":96},"biblio":{"volume":"90","issue":"3","first_page":"753","last_page":"799"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9987999796867371,"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":0.9987999796867371,"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":0.9930999875068665,"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/T10963","display_name":"Advanced Optimization Algorithms Research","score":0.9847999811172485,"subfield":{"id":"https://openalex.org/subfields/2612","display_name":"Numerical Analysis"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.9150004982948303},{"id":"https://openalex.org/keywords/rounding","display_name":"Rounding","score":0.8091060519218445},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.6287814378738403},{"id":"https://openalex.org/keywords/descent","display_name":"Descent (aeronautics)","score":0.5635937452316284},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"Stochastic gradient descent","score":0.5390623807907104},{"id":"https://openalex.org/keywords/fixed-point","display_name":"Fixed point","score":0.4967051148414612},{"id":"https://openalex.org/keywords/inequality","display_name":"Inequality","score":0.4889172613620758},{"id":"https://openalex.org/keywords/applied-mathematics","display_name":"Applied mathematics","score":0.47725412249565125},{"id":"https://openalex.org/keywords/rate-of-convergence","display_name":"Rate of convergence","score":0.42329761385917664},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.40229305624961853},{"id":"https://openalex.org/keywords/mathematical-analysis","display_name":"Mathematical analysis","score":0.2776705324649811},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.05276140570640564}],"concepts":[{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.9150004982948303},{"id":"https://openalex.org/C136625980","wikidata":"https://www.wikidata.org/wiki/Q663208","display_name":"Rounding","level":2,"score":0.8091060519218445},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.6287814378738403},{"id":"https://openalex.org/C2776637919","wikidata":"https://www.wikidata.org/wiki/Q624380","display_name":"Descent (aeronautics)","level":2,"score":0.5635937452316284},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.5390623807907104},{"id":"https://openalex.org/C61445026","wikidata":"https://www.wikidata.org/wiki/Q217608","display_name":"Fixed point","level":2,"score":0.4967051148414612},{"id":"https://openalex.org/C45555294","wikidata":"https://www.wikidata.org/wiki/Q28113351","display_name":"Inequality","level":2,"score":0.4889172613620758},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.47725412249565125},{"id":"https://openalex.org/C57869625","wikidata":"https://www.wikidata.org/wiki/Q1783502","display_name":"Rate of convergence","level":3,"score":0.42329761385917664},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.40229305624961853},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.2776705324649811},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.05276140570640564},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","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},{"id":"https://openalex.org/C146978453","wikidata":"https://www.wikidata.org/wiki/Q3798668","display_name":"Aerospace engineering","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/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","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/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"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":4,"locations":[{"id":"doi:10.1007/s10589-025-00656-1","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s10589-025-00656-1","pdf_url":"https://link.springer.com/content/pdf/10.1007/s10589-025-00656-1.pdf","source":{"id":"https://openalex.org/S39243548","display_name":"Computational Optimization and Applications","issn_l":"0926-6003","issn":["0926-6003","1573-2894"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Computational Optimization and Applications","raw_type":"journal-article"},{"id":"pmh:oai:pure.tue.nl:openaire/9387b0ce-6d8e-4bd2-8c6e-08c49ea4f02c","is_oa":true,"landing_page_url":"https://research.tue.nl/en/publications/9387b0ce-6d8e-4bd2-8c6e-08c49ea4f02c","pdf_url":null,"source":{"id":"https://openalex.org/S4406922641","display_name":"TU/e Research Portal","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Xia, L, Massei, S & Hochstenbach, M E 2025, 'On the convergence of the gradient descent method with stochastic fixed-point rounding errors under the Polyak\u2013\u0141ojasiewicz inequality', Computational Optimization and Applications, vol. 90, no. 3, pp. 753-799. https://doi.org/10.1007/s10589-025-00656-1","raw_type":"info:eu-repo/semantics/article"},{"id":"pmh:oai:RePEc:spr:coopap:v:90:y:2025:i:3:d:10.1007_s10589-025-00656-1","is_oa":false,"landing_page_url":"http://link.springer.com/10.1007/s10589-025-00656-1","pdf_url":null,"source":{"id":"https://openalex.org/S4306401271","display_name":"RePEc: Research Papers in Economics","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I77793887","host_organization_name":"Federal Reserve Bank of St. Louis","host_organization_lineage":["https://openalex.org/I77793887"],"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":"article"},{"id":"pmh:oai:arpi.unipi.it:11568/1329375","is_oa":false,"landing_page_url":"https://hdl.handle.net/11568/1329375","pdf_url":null,"source":{"id":"https://openalex.org/S4377196265","display_name":"CINECA IRIS Institutial research information system (University of Pisa)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I108290504","host_organization_name":"University of Pisa","host_organization_lineage":["https://openalex.org/I108290504"],"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":"info:eu-repo/semantics/article"}],"best_oa_location":{"id":"doi:10.1007/s10589-025-00656-1","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s10589-025-00656-1","pdf_url":"https://link.springer.com/content/pdf/10.1007/s10589-025-00656-1.pdf","source":{"id":"https://openalex.org/S39243548","display_name":"Computational Optimization and Applications","issn_l":"0926-6003","issn":["0926-6003","1573-2894"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Computational Optimization and Applications","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G2998642118","display_name":"Arrowhead Tools for Engineering of Digitalisation Solutions","funder_award_id":"826452","funder_id":"https://openalex.org/F4320338475","funder_display_name":"H2020 LEIT Information and Communication Technologies"}],"funders":[{"id":"https://openalex.org/F4320338475","display_name":"H2020 LEIT Information and Communication Technologies","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4407367788.pdf","grobid_xml":"https://content.openalex.org/works/W4407367788.grobid-xml"},"referenced_works_count":23,"referenced_works":["https://openalex.org/W1817239747","https://openalex.org/W1971942712","https://openalex.org/W1987217557","https://openalex.org/W1988720110","https://openalex.org/W1994961181","https://openalex.org/W2018435387","https://openalex.org/W2020804487","https://openalex.org/W2268760178","https://openalex.org/W2593239304","https://openalex.org/W2611105550","https://openalex.org/W2736068458","https://openalex.org/W2938736712","https://openalex.org/W2963248893","https://openalex.org/W3023393331","https://openalex.org/W3141595720","https://openalex.org/W3176058568","https://openalex.org/W3186819059","https://openalex.org/W3211635928","https://openalex.org/W4226038297","https://openalex.org/W4250589301","https://openalex.org/W4390002681","https://openalex.org/W6606667432","https://openalex.org/W6814268980"],"related_works":["https://openalex.org/W4220780102","https://openalex.org/W2410881844","https://openalex.org/W3196334750","https://openalex.org/W4308900592","https://openalex.org/W3194588123","https://openalex.org/W2964170259","https://openalex.org/W3002546633","https://openalex.org/W4206119629","https://openalex.org/W2765682467","https://openalex.org/W4382937879"],"abstract_inverted_index":{"Abstract":[0],"In":[1],"the":[2,22,25,32,41,48,51,58,82,88,102,109,123],"training":[3],"of":[4,24,34,44,50,125],"neural":[5],"networks":[6],"with":[7],"low-precision":[8,134],"computation":[9],"and":[10,86],"fixed-point":[11,135],"arithmetic,":[12],"rounding":[13,37,70,127],"errors":[14,46],"often":[15],"cause":[16],"stagnation":[17],"or":[18],"are":[19],"detrimental":[20],"to":[21,39],"convergence":[23,49,103],"optimizers.":[26],"This":[27],"study":[28],"provides":[29],"insights":[30],"into":[31],"choice":[33],"appropriate":[35],"stochastic":[36,69,114],"strategies":[38,128],"mitigate":[40],"adverse":[42],"impact":[43],"roundoff":[45,90],"on":[47,101],"gradient":[52,84],"descent":[53,94],"method,":[54],"for":[55],"problems":[56],"satisfying":[57],"Polyak\u2013\u0141ojasiewicz":[59],"inequality.":[60],"Within":[61],"this":[62],"context,":[63],"we":[64,97],"show":[65],"that":[66,105],"a":[67,93,99],"biased":[68],"strategy":[71],"may":[72],"be":[73],"even":[74],"beneficial":[75],"in":[76,92],"so":[77],"far":[78],"as":[79],"it":[80],"eliminates":[81],"vanishing":[83],"problem":[85],"forces":[87],"expected":[89],"error":[91],"direction.":[95],"Furthermore,":[96],"obtain":[98],"bound":[100],"rate":[104],"is":[106,119],"stricter":[107],"than":[108],"one":[110],"achieved":[111],"by":[112,121],"unbiased":[113],"rounding.":[115],"The":[116],"theoretical":[117],"analysis":[118],"validated":[120],"comparing":[122],"performances":[124],"various":[126],"when":[129],"optimizing":[130],"several":[131],"examples":[132],"using":[133],"arithmetic.":[136]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2026-06-13T06:13:01.061226","created_date":"2025-10-10T00:00:00"}