{"id":"https://openalex.org/W2963724148","doi":"https://doi.org/10.5555/2789272.2912107","title":"On the inductive bias of dropout","display_name":"On the inductive bias of dropout","publication_year":2015,"publication_date":"2015-01-01","ids":{"openalex":"https://openalex.org/W2963724148","doi":"https://doi.org/10.5555/2789272.2912107","mag":"2963724148"},"language":"en","primary_location":{"id":"mag:2963724148","is_oa":false,"landing_page_url":"http://jmlr.csail.mit.edu/papers/volume16/helmbold15a/helmbold15a.pdf","pdf_url":null,"source":{"id":"https://openalex.org/S118988714","display_name":"Journal of Machine Learning Research","issn_l":"1532-4435","issn":["1532-4435","1533-7928"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310315718","host_organization_name":"The MIT Press","host_organization_lineage":["https://openalex.org/P4310315718"],"host_organization_lineage_names":["The MIT Press"],"type":"journal"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":"Journal of Machine Learning Research","raw_type":null},"type":"article","indexed_in":[],"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/A5014740431","display_name":"David P. Helmbold","orcid":null},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"David P. Helmbold","raw_affiliation_strings":["Department of Computer Science, University of California Santa Cruz, Santa Cruz, CA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science, University of California Santa Cruz, Santa Cruz, CA","institution_ids":["https://openalex.org/I185103710"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5084487318","display_name":"Philip M. Long","orcid":"https://orcid.org/0000-0002-1010-6197"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Philip M. Long","raw_affiliation_strings":["Microsoft,Sunnyvale,CA"],"raw_orcid":"https://orcid.org/0000-0002-1010-6197","affiliations":[{"raw_affiliation_string":"Microsoft,Sunnyvale,CA","institution_ids":["https://openalex.org/I1290206253"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5014740431"],"corresponding_institution_ids":["https://openalex.org/I185103710"],"apc_list":null,"apc_paid":null,"fwci":7.1221,"has_fulltext":false,"cited_by_count":35,"citation_normalized_percentile":{"value":0.97081834,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":90,"max":99},"biblio":{"volume":"16","issue":"1","first_page":"3403","last_page":"3454"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9998999834060669,"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":0.9998999834060669,"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":0.9998000264167786,"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":0.9991999864578247,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/regularization","display_name":"Regularization (linguistics)","score":0.764896035194397},{"id":"https://openalex.org/keywords/bounded-function","display_name":"Bounded function","score":0.5144944190979004},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.46991172432899475},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.46692463755607605},{"id":"https://openalex.org/keywords/regular-polygon","display_name":"Regular polygon","score":0.46299347281455994},{"id":"https://openalex.org/keywords/dropout","display_name":"Dropout (neural networks)","score":0.444345086812973},{"id":"https://openalex.org/keywords/convex-function","display_name":"Convex function","score":0.44129911065101624},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.4229121208190918},{"id":"https://openalex.org/keywords/applied-mathematics","display_name":"Applied mathematics","score":0.38249075412750244},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.33154886960983276},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.22610238194465637},{"id":"https://openalex.org/keywords/mathematical-analysis","display_name":"Mathematical analysis","score":0.0955875813961029}],"concepts":[{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.764896035194397},{"id":"https://openalex.org/C34388435","wikidata":"https://www.wikidata.org/wiki/Q2267362","display_name":"Bounded function","level":2,"score":0.5144944190979004},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.46991172432899475},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.46692463755607605},{"id":"https://openalex.org/C112680207","wikidata":"https://www.wikidata.org/wiki/Q714886","display_name":"Regular polygon","level":2,"score":0.46299347281455994},{"id":"https://openalex.org/C2776145597","wikidata":"https://www.wikidata.org/wiki/Q25339462","display_name":"Dropout (neural networks)","level":2,"score":0.444345086812973},{"id":"https://openalex.org/C145446738","wikidata":"https://www.wikidata.org/wiki/Q319913","display_name":"Convex function","level":3,"score":0.44129911065101624},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.4229121208190918},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.38249075412750244},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33154886960983276},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.22610238194465637},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0955875813961029},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:2963724148","is_oa":false,"landing_page_url":"http://jmlr.csail.mit.edu/papers/volume16/helmbold15a/helmbold15a.pdf","pdf_url":null,"source":{"id":"https://openalex.org/S118988714","display_name":"Journal of Machine Learning Research","issn_l":"1532-4435","issn":["1532-4435","1533-7928"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310315718","host_organization_name":"The MIT Press","host_organization_lineage":["https://openalex.org/P4310315718"],"host_organization_lineage_names":["The MIT Press"],"type":"journal"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":"Journal of Machine Learning Research","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/8","score":0.6899999976158142,"display_name":"Decent work and economic growth"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":23,"referenced_works":["https://openalex.org/W4919037","https://openalex.org/W35527955","https://openalex.org/W620038288","https://openalex.org/W1576619248","https://openalex.org/W1904365287","https://openalex.org/W1982032418","https://openalex.org/W1984541135","https://openalex.org/W2005306124","https://openalex.org/W2023163512","https://openalex.org/W2035446310","https://openalex.org/W2062227835","https://openalex.org/W2095705004","https://openalex.org/W2129068307","https://openalex.org/W2135754715","https://openalex.org/W2144459111","https://openalex.org/W2146502635","https://openalex.org/W2152722485","https://openalex.org/W2164370980","https://openalex.org/W2294567968","https://openalex.org/W2490901831","https://openalex.org/W2531891978","https://openalex.org/W2612090632","https://openalex.org/W2616625443"],"related_works":["https://openalex.org/W2152722485","https://openalex.org/W2095705004","https://openalex.org/W1904365287","https://openalex.org/W2294567968","https://openalex.org/W4919037","https://openalex.org/W1746889935","https://openalex.org/W2964059111","https://openalex.org/W35527955","https://openalex.org/W2803183915","https://openalex.org/W3118608800","https://openalex.org/W2962857907","https://openalex.org/W2164370980","https://openalex.org/W2111406701","https://openalex.org/W2963649970","https://openalex.org/W2144459111","https://openalex.org/W2939195450","https://openalex.org/W2144513243","https://openalex.org/W2097117768","https://openalex.org/W2963918617","https://openalex.org/W3034384909"],"abstract_inverted_index":{"Dropout":[0],"is":[1,21,75,132,146],"a":[2,47,60,85,148],"simple":[3],"but":[4],"effective":[5],"technique":[6],"for":[7,215],"learning":[8],"in":[9,72],"neural":[10],"networks":[11],"and":[12,30,99,118,191,206],"other":[13],"settings.":[14],"A":[15],"sound":[16],"theoretical":[17],"understanding":[18],"of":[19,44,138,163,168,204],"dropout":[20,26,45,107,122,139,154,186,205],"needed":[22],"to":[23,32,63,94,152],"determine":[24],"when":[25,80,89,100,164],"should":[27],"be":[28,110,127],"applied":[29],"how":[31,200],"use":[33],"it":[34,101],"most":[35],"effectively.":[36],"In":[37],"this":[38],"paper":[39],"we":[40,159],"continue":[41],"the":[42,64,68,81,90,97,106,121,136,161,201],"exploration":[43],"as":[46,112],"regularizer":[48],"pioneered":[49],"by":[50],"Wager":[51],"et":[52],"al.":[53],"We":[54,77,176,210],"focus":[55],"on":[56],"linear":[57],"classification":[58],"where":[59],"convex":[61,143,149],"proxy":[62,145],"misclassification":[65],"loss":[66,70,144],"(i.e.":[67],"logistic":[69,73],"used":[71],"regression)":[74],"minimized.":[76],"show:":[78],"\u2022":[79,88,104,119],"dropout-regularized":[82],"criterion":[83],"has":[84],"unique":[86],"minimizer,":[87],"dropout-regularization":[91],"penalty":[92,124],"goes":[93],"infinity":[95],"with":[96,141,156,173,185],"weights,":[98],"remains":[102],"bounded,":[103],"that":[105,120,180],"regularization":[108,123,140,155,187,208],"can":[109],"non-monotonic":[111],"individual":[113],"weights":[114],"increase":[115],"from":[116],"0,":[117],"may":[125],"not":[126],"convex.":[128],"This":[129],"last":[130],"point":[131],"particularly":[133],"surprising":[134],"because":[135],"combination":[137],"any":[142],"always":[147],"function.\r\n\r\nIn":[150],"order":[151],"contrast":[153],"L2":[157,189,207],"regularization,":[158,190],"formalize":[160],"notion":[162],"different":[165,174],"random":[166],"sources":[167,195],"data":[169],"are":[170,181],"more":[171,183],"compatible":[172,184],"regularizers.":[175],"then":[177],"exhibit":[178],"distributions":[179],"provably":[182],"than":[188],"vice":[192],"versa.":[193],"These":[194],"provide":[196,211],"additional":[197],"insight":[198],"into":[199],"inductive":[202],"biases":[203],"differ.":[209],"some":[212],"similar":[213],"results":[214],"L1":[216],"regularization.":[217]},"counts_by_year":[{"year":2021,"cited_by_count":5},{"year":2020,"cited_by_count":7},{"year":2019,"cited_by_count":7},{"year":2018,"cited_by_count":5},{"year":2017,"cited_by_count":5},{"year":2016,"cited_by_count":5},{"year":2015,"cited_by_count":1}],"updated_date":"2026-04-28T14:05:53.105641","created_date":"2025-10-10T00:00:00"}