{"id":"https://openalex.org/W4372260056","doi":"https://doi.org/10.1109/icassp49357.2023.10095934","title":"On Weighted Cross-Entropy for Label-Imbalanced Separable Data: An Algorithmic-Stability Study","display_name":"On Weighted Cross-Entropy for Label-Imbalanced Separable Data: An Algorithmic-Stability Study","publication_year":2023,"publication_date":"2023-05-05","ids":{"openalex":"https://openalex.org/W4372260056","doi":"https://doi.org/10.1109/icassp49357.2023.10095934"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49357.2023.10095934","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10095934","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 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/A5053095531","display_name":"Puneesh Deora","orcid":null},"institutions":[{"id":"https://openalex.org/I141945490","display_name":"University of British Columbia","ror":"https://ror.org/03rmrcq20","country_code":"CA","type":"education","lineage":["https://openalex.org/I141945490"]}],"countries":["CA"],"is_corresponding":true,"raw_author_name":"Puneesh Deora","raw_affiliation_strings":["University of British Columbia"],"affiliations":[{"raw_affiliation_string":"University of British Columbia","institution_ids":["https://openalex.org/I141945490"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5024812488","display_name":"Christos Thrampoulidis","orcid":"https://orcid.org/0000-0001-9053-9365"},"institutions":[{"id":"https://openalex.org/I141945490","display_name":"University of British Columbia","ror":"https://ror.org/03rmrcq20","country_code":"CA","type":"education","lineage":["https://openalex.org/I141945490"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Christos Thrampoulidis","raw_affiliation_strings":["University of British Columbia"],"affiliations":[{"raw_affiliation_string":"University of British Columbia","institution_ids":["https://openalex.org/I141945490"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5053095531"],"corresponding_institution_ids":["https://openalex.org/I141945490"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.03749489,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9990000128746033,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9990000128746033,"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":0.9980000257492065,"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":0.9976999759674072,"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/computer-science","display_name":"Computer science","score":0.5348763465881348},{"id":"https://openalex.org/keywords/early-stopping","display_name":"Early stopping","score":0.5319336652755737},{"id":"https://openalex.org/keywords/generalization-error","display_name":"Generalization error","score":0.4987678527832031},{"id":"https://openalex.org/keywords/logarithm","display_name":"Logarithm","score":0.49548232555389404},{"id":"https://openalex.org/keywords/cross-entropy","display_name":"Cross entropy","score":0.4822719693183899},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"Gradient descent","score":0.4684355854988098},{"id":"https://openalex.org/keywords/conjecture","display_name":"Conjecture","score":0.44543248414993286},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.43411922454833984},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.4336013197898865},{"id":"https://openalex.org/keywords/regularization","display_name":"Regularization (linguistics)","score":0.4217957854270935},{"id":"https://openalex.org/keywords/applied-mathematics","display_name":"Applied mathematics","score":0.39479348063468933},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.36964234709739685},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3418040871620178},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.30121850967407227},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.18439114093780518},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.18114179372787476},{"id":"https://openalex.org/keywords/discrete-mathematics","display_name":"Discrete mathematics","score":0.17500609159469604}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5348763465881348},{"id":"https://openalex.org/C5465570","wikidata":"https://www.wikidata.org/wiki/Q5326898","display_name":"Early stopping","level":3,"score":0.5319336652755737},{"id":"https://openalex.org/C117765406","wikidata":"https://www.wikidata.org/wiki/Q5362437","display_name":"Generalization error","level":3,"score":0.4987678527832031},{"id":"https://openalex.org/C39927690","wikidata":"https://www.wikidata.org/wiki/Q11197","display_name":"Logarithm","level":2,"score":0.49548232555389404},{"id":"https://openalex.org/C167981619","wikidata":"https://www.wikidata.org/wiki/Q1685498","display_name":"Cross entropy","level":3,"score":0.4822719693183899},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.4684355854988098},{"id":"https://openalex.org/C2780990831","wikidata":"https://www.wikidata.org/wiki/Q319141","display_name":"Conjecture","level":2,"score":0.44543248414993286},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.43411922454833984},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4336013197898865},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.4217957854270935},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.39479348063468933},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.36964234709739685},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3418040871620178},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.30121850967407227},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.18439114093780518},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.18114179372787476},{"id":"https://openalex.org/C118615104","wikidata":"https://www.wikidata.org/wiki/Q121416","display_name":"Discrete mathematics","level":1,"score":0.17500609159469604},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49357.2023.10095934","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10095934","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":40,"referenced_works":["https://openalex.org/W1811750039","https://openalex.org/W2790253170","https://openalex.org/W2921393178","https://openalex.org/W2965772785","https://openalex.org/W2970941190","https://openalex.org/W2984133899","https://openalex.org/W2984169168","https://openalex.org/W2997580736","https://openalex.org/W3018252856","https://openalex.org/W3021336872","https://openalex.org/W3034851139","https://openalex.org/W3035703482","https://openalex.org/W3122855191","https://openalex.org/W3137695714","https://openalex.org/W4221163054","https://openalex.org/W4287285260","https://openalex.org/W4287748599","https://openalex.org/W4288413101","https://openalex.org/W4299785371","https://openalex.org/W4301881812","https://openalex.org/W6638214083","https://openalex.org/W6745276634","https://openalex.org/W6747679287","https://openalex.org/W6748742374","https://openalex.org/W6755424776","https://openalex.org/W6760201928","https://openalex.org/W6760498417","https://openalex.org/W6763485134","https://openalex.org/W6764733053","https://openalex.org/W6766887723","https://openalex.org/W6769472860","https://openalex.org/W6772178927","https://openalex.org/W6774616501","https://openalex.org/W6776699476","https://openalex.org/W6778993914","https://openalex.org/W6780430711","https://openalex.org/W6780793664","https://openalex.org/W6791656692","https://openalex.org/W6809714182","https://openalex.org/W7034108470"],"related_works":["https://openalex.org/W2136627871","https://openalex.org/W2153743018","https://openalex.org/W1535041666","https://openalex.org/W4206677161","https://openalex.org/W1494709439","https://openalex.org/W69760452","https://openalex.org/W2117037448","https://openalex.org/W4380225034","https://openalex.org/W2134396573","https://openalex.org/W4309434778"],"abstract_inverted_index":{"Implicit":[0],"bias":[1],"theory":[2],"characterizes":[3],"notions":[4],"of":[5,35,44,159,176,190],"simplicity":[6],"in":[7],"the":[8,63,69,74,89,101,125,135,148,151,156,170,173,177,181],"weights":[9],"learned":[10],"by":[11,67,114],"gradient":[12,45],"descent":[13],"when":[14],"training":[15,21,81,130],"without":[16],"explicit":[17],"regularization":[18],"beyond":[19],"zero":[20],"error,":[22],"and":[23,154],"has":[24],"served":[25],"as":[26,94,100,172],"a":[27,187],"cornerstone":[28],"result":[29],"for":[30,78,193],"theoretically":[31],"justifying":[32],"good":[33,188],"generalization":[34,55,64,109],"interpolating":[36],"models.":[37],"However,":[38],"its":[39,48],"asymptotic":[40],"nature":[41],"(in":[42],"number":[43],"steps)":[46],"limits":[47],"practical":[49],"relevance.":[50],"This":[51],"motivates":[52],"developing":[53],"finite-time":[54],"bounds.":[56],"Specifically,":[57],"recent":[58],"works":[59],"have":[60],"proposed":[61],"bounding":[62],"error":[65,175,192],"indirectly":[66],"controlling":[68],"corresponding":[70],"test":[71,91,102,157,182],"loss":[72,92,158,183],"via":[73],"algorithmic-stability":[75],"framework.":[76],"Concretely,":[77],"cross-entropy":[79],"(CE)":[80],"on":[82],"separable":[83],"balanced":[84,174,191],"data,":[85],"they":[86],"show":[87,165],"that":[88,118],"CE":[90,120],"decays":[93],"fast":[95],"(up":[96],"to":[97,150],"logarithmic":[98],"factors)":[99],"error.":[103],"In":[104],"this":[105,140,144,166,200],"paper,":[106],"we":[107,132,146,164],"study":[108],"under":[110,195],"label":[111],"imbalances.":[112,196],"Motivated":[113],"our":[115],"empirical":[116,203],"observation":[117],"weighted":[119],"(wCE)":[121],"can":[122,138],"significantly":[123],"outperform":[124],"max-margin":[126,178],"classifier":[127],"at":[128],"early":[129],"phases,":[131],"ask":[133],"whether":[134],"stability":[136],"framework":[137],"prove":[139],"early-stopping":[141],"result.":[142],"To":[143],"end,":[145],"extend":[147],"analysis":[149],"imbalanced":[152],"setting":[153],"bound":[155,167],"wCE.":[160],"For":[161],"Gaussian":[162],"mixtures,":[163],"is":[168],"orderwise":[169],"same":[171],"classifier,":[179],"suggesting":[180],"might":[184],"not":[185],"be":[186],"proxy":[189],"wCE":[194],"We":[197],"further":[198],"support":[199],"conjecture":[201],"with":[202],"results.":[204]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}