{"id":"https://openalex.org/W7225099192","doi":"https://doi.org/10.5281/zenodo.23873893","title":"Gradient Descent Finds the Maximum-Margin Classifier, Slowly: Ten Million Iterations of Unregularised Logistic Regression on Two Separable UCI Data Sets, a Norm\u2013Loss Bound, and Early Stopping","display_name":"Gradient Descent Finds the Maximum-Margin Classifier, Slowly: Ten Million Iterations of Unregularised Logistic Regression on Two Separable UCI Data Sets, a Norm\u2013Loss Bound, and Early Stopping","publication_year":2026,"publication_date":"2026-11-02","ids":{"openalex":"https://openalex.org/W7225099192","doi":"https://doi.org/10.5281/zenodo.23873893"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.23873893","doi":"https://doi.org/10.5281/zenodo.23873893","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23873893","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.23873893","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5159002730","display_name":"Manoj Kumar Aguru","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405266258","display_name":"Chaitanya","ror":"https://ror.org/01m1bpg27","country_code":"IN","type":"education","lineage":["https://openalex.org/I4405266258"]}],"countries":["IN"],"is_corresponding":true,"raw_author_name":"Aguru, Manoj Kumar","raw_affiliation_strings":["Former Officer, Chaitanya Godavari Grameena Bank (CGGB)"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Former Officer, Chaitanya Godavari Grameena Bank (CGGB)","institution_ids":["https://openalex.org/I4405266258"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5159002730"],"corresponding_institution_ids":["https://openalex.org/I4405266258"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"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":0.9682999849319458,"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":0.9682999849319458,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.015599999576807022,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.006500000134110451,"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/gradient-descent","display_name":"gradient descent","score":0.7649999856948853},{"id":"https://openalex.org/keywords/maximum-margin-classifier","display_name":"maximum margin classifier","score":0.8040000200271606},{"id":"https://openalex.org/keywords/linear-separability","display_name":"linear separability","score":0.7960000038146973},{"id":"https://openalex.org/keywords/early-stopping","display_name":"early stopping","score":0.7519999742507935},{"id":"https://openalex.org/keywords/karush-kuhn-tucker-conditions","display_name":"Karush-Kuhn-Tucker conditions","score":0.7870000004768372},{"id":"https://openalex.org/keywords/convergence-rate","display_name":"convergence rate","score":0.47999998927116394},{"id":"https://openalex.org/keywords/sonar-classification","display_name":"sonar classification","score":0.7070000171661377}],"concepts":[{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.7203999757766724},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.54830002784729},{"id":"https://openalex.org/C70710897","wikidata":"https://www.wikidata.org/wiki/Q680081","display_name":"Separable space","level":2,"score":0.5432999730110168},{"id":"https://openalex.org/C140479938","wikidata":"https://www.wikidata.org/wiki/Q5254619","display_name":"Iterated function","level":2,"score":0.5217000246047974},{"id":"https://openalex.org/C191795146","wikidata":"https://www.wikidata.org/wiki/Q3878446","display_name":"Norm (philosophy)","level":2,"score":0.4853000044822693},{"id":"https://openalex.org/C151956035","wikidata":"https://www.wikidata.org/wiki/Q1132755","display_name":"Logistic regression","level":2,"score":0.48100000619888306},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.42419999837875366},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.4020000100135803},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.3598000109195709},{"id":"https://openalex.org/C122041747","wikidata":"https://www.wikidata.org/wiki/Q838611","display_name":"Ball (mathematics)","level":2,"score":0.3456000089645386},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.3440000116825104},{"id":"https://openalex.org/C122048520","wikidata":"https://www.wikidata.org/wiki/Q2913954","display_name":"Percentile","level":2,"score":0.2976999878883362},{"id":"https://openalex.org/C12267149","wikidata":"https://www.wikidata.org/wiki/Q282453","display_name":"Support vector machine","level":2,"score":0.2937000095844269},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.29019999504089355},{"id":"https://openalex.org/C191948623","wikidata":"https://www.wikidata.org/wiki/Q1307800","display_name":"Unit sphere","level":2,"score":0.28459998965263367},{"id":"https://openalex.org/C48921125","wikidata":"https://www.wikidata.org/wiki/Q10861030","display_name":"Linear regression","level":2,"score":0.2802000045776367},{"id":"https://openalex.org/C157553263","wikidata":"https://www.wikidata.org/wiki/Q5168004","display_name":"Coordinate descent","level":2,"score":0.27079999446868896},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.26969999074935913},{"id":"https://openalex.org/C83546350","wikidata":"https://www.wikidata.org/wiki/Q1139051","display_name":"Regression","level":2,"score":0.265500009059906},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2630999982357025},{"id":"https://openalex.org/C180802074","wikidata":"https://www.wikidata.org/wiki/Q1052379","display_name":"Logistic function","level":2,"score":0.2603999972343445},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.2563000023365021},{"id":"https://openalex.org/C21080849","wikidata":"https://www.wikidata.org/wiki/Q13611879","display_name":"Data point","level":2,"score":0.25519999861717224},{"id":"https://openalex.org/C555745239","wikidata":"https://www.wikidata.org/wiki/Q133220","display_name":"Sonar","level":2,"score":0.25290000438690186}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.23873893","doi":"https://doi.org/10.5281/zenodo.23873893","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23873893","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.5281/zenodo.23873894","doi":"https://doi.org/10.5281/zenodo.23873894","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23873894","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.5281/zenodo.23873893","doi":"https://doi.org/10.5281/zenodo.23873893","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23873893","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2766965791","https://openalex.org/W7223069161","https://openalex.org/W7220697009","https://openalex.org/W2949247311","https://openalex.org/W4416125235","https://openalex.org/W7129085708","https://openalex.org/W2103718066","https://openalex.org/W4392224005","https://openalex.org/W7216454620","https://openalex.org/W7197047026"],"abstract_inverted_index":{"When":[0],"the":[1,30,91,117,127,132,149,161,164,179,203,214,221,226,251],"training":[2,103,190],"data":[3,61,145,165],"can":[4],"be":[5,120],"split":[6],"perfectly":[7],"by":[8,100,153],"a":[9,113,173,197],"hyperplane,":[10],"unregularised":[11],"logistic":[12,93],"regression":[13],"has":[14,242],"no":[15],"finite":[16],"solution,":[17],"yet":[18],"gradient":[19,236],"descent":[20,237],"on":[21,48,143,245,259],"it":[22],"is":[23,40,105,131,229],"known":[24],"to":[25,29],"converge":[26],"in":[27],"direction":[28],"hard-margin":[31],"support":[32],"vector":[33,111],"machine.":[34],"The":[35,217],"theory":[36],"says":[37],"this":[38],"convergence":[39],"logarithmically":[41],"slow.":[42],"This":[43],"paper":[44],"measures":[45],"how":[46],"slow":[47,227],"real":[49],"data.":[50],"A":[51],"linear":[52],"programme":[53],"shows":[54],"that":[55],"two":[56,87],"of":[57,176,200,210,224],"four":[58],"standard":[59],"UCI":[60],"sets":[62],"are":[63,83],"linearly":[64],"separable":[65],"after":[66,192],"standardisation:":[67],"breast":[68,183,246],"cancer":[69,184,247],"Wisconsin":[70],"diagnostic":[71],"(569":[72],"cases,":[73],"30":[74],"features)":[75],"and":[76,81,108,147,171,202,253],"sonar":[77,158,260],"(208,":[78],"60);":[79],"banknote":[80],"ionosphere":[82],"not.":[84],"I":[85,136],"prove":[86],"elementary":[88],"facts:":[89],"if":[90],"average":[92],"loss":[94],"falls":[95],"below":[96],"ln":[97],"2":[98],"divided":[99],"n,":[101],"every":[102,109],"point":[104],"correctly":[106],"classified,":[107],"weight":[110],"obeys":[112],"norm\u2013loss":[114],"bound,":[115],"so":[116],"norm":[118,204],"must":[119],"at":[121,206,239,255],"least":[122],"(1/gamma)":[123],"ln(ln":[124],"2/(n":[125],"times":[126],"loss))":[128],"where":[129],"gamma":[130],"maximum":[133],"geometric":[134],"margin.":[135],"then":[137],"run":[138],"ten":[139,193],"million":[140,169,194],"gradient-descent":[141],"steps":[142,170,241],"each":[144],"set":[146],"certify":[148],"exact":[150],"max-margin":[151],"solution":[152],"its":[154],"KKT":[155],"conditions.":[156],"On":[157,182],"(margin":[159,185],"0.0196)":[160],"iterates":[162],"separate":[163],"from":[166],"about":[167],"one":[168],"reach":[172],"cosine":[174,198],"gap":[175,199],"0.00226":[177],"with":[178,196,233],"SVM":[180],"direction.":[181],"0.0014)":[186],"they":[187],"still":[188],"misclassify":[189],"points":[191],"steps,":[195,257],"0.060,":[201],"grows":[205],"12.5":[207],"per":[208],"unit":[209],"log":[211],"t":[212],"against":[213,248,261],"asymptotic":[215],"718.":[216],"small":[218],"margin":[219],"explains":[220],"slowness.":[222],"Out":[223],"sample,":[225],"road":[228],"not":[230],"worth":[231],"travelling:":[232],"five-fold":[234],"cross-validation,":[235],"stopped":[238,254],"1,000":[240],"error":[243],"0.0211":[244],"0.0475":[249],"for":[250],"SVM,":[252],"10":[256],"0.2260":[258],"0.2837.":[262]},"counts_by_year":[],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2026-11-03T00:00:00"}