{"id":"https://openalex.org/W4309315585","doi":"https://doi.org/10.3390/make4040051","title":"Evidence-Based Regularization for Neural Networks","display_name":"Evidence-Based Regularization for Neural Networks","publication_year":2022,"publication_date":"2022-11-15","ids":{"openalex":"https://openalex.org/W4309315585","doi":"https://doi.org/10.3390/make4040051"},"language":"en","primary_location":{"id":"doi:10.3390/make4040051","is_oa":true,"landing_page_url":"https://doi.org/10.3390/make4040051","pdf_url":"https://www.mdpi.com/2504-4990/4/4/51/pdf?version=1668501014","source":{"id":"https://openalex.org/S4210213891","display_name":"Machine Learning and Knowledge Extraction","issn_l":"2504-4990","issn":["2504-4990"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310310987","host_organization_name":"Multidisciplinary Digital Publishing Institute","host_organization_lineage":["https://openalex.org/P4310310987"],"host_organization_lineage_names":["Multidisciplinary Digital Publishing Institute"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Machine Learning and Knowledge Extraction","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.mdpi.com/2504-4990/4/4/51/pdf?version=1668501014","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5036837747","display_name":"Giuseppe Nuti","orcid":"https://orcid.org/0000-0001-8864-318X"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Giuseppe Nuti","raw_affiliation_strings":["UBS Investment Bank, 1285 Avenue of the Americas, New York, NY 10019, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UBS Investment Bank, 1285 Avenue of the Americas, New York, NY 10019, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109413521","display_name":"Andreea-Ingrid Cross","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Andreea-Ingrid Cross","raw_affiliation_strings":["UBS Investment Bank, 5 Broadgate, London EC2M 2QS, UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UBS Investment Bank, 5 Broadgate, London EC2M 2QS, UK","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5048527077","display_name":"Philipp Rindler","orcid":null},"institutions":[{"id":"https://openalex.org/I131328143","display_name":"Union Bank of Switzerland","ror":"https://ror.org/01ekk8h85","country_code":"CH","type":"funder","lineage":["https://openalex.org/I131328143"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Philipp Rindler","raw_affiliation_strings":["UBS Business Solutions AG CH, Europaallee 21, 8004 Zurich, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UBS Business Solutions AG CH, Europaallee 21, 8004 Zurich, Switzerland","institution_ids":["https://openalex.org/I131328143"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5036837747"],"corresponding_institution_ids":[],"apc_list":{"value":1400,"currency":"CHF","value_usd":1515},"apc_paid":{"value":1400,"currency":"CHF","value_usd":1515},"fwci":0.8323,"has_fulltext":true,"cited_by_count":6,"citation_normalized_percentile":{"value":0.78140947,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":97},"biblio":{"volume":"4","issue":"4","first_page":"1011","last_page":"1023"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9991000294685364,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9991000294685364,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9990000128746033,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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"}}],"keywords":[{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.6357229948043823},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6284238696098328},{"id":"https://openalex.org/keywords/hyperplane","display_name":"Hyperplane","score":0.6068750619888306},{"id":"https://openalex.org/keywords/maxima-and-minima","display_name":"Maxima and minima","score":0.5886716246604919},{"id":"https://openalex.org/keywords/regularization","display_name":"Regularization (linguistics)","score":0.587266206741333},{"id":"https://openalex.org/keywords/a-priori-and-a-posteriori","display_name":"A priori and a posteriori","score":0.5627576112747192},{"id":"https://openalex.org/keywords/data-point","display_name":"Data point","score":0.48823121190071106},{"id":"https://openalex.org/keywords/gaussian","display_name":"Gaussian","score":0.44410115480422974},{"id":"https://openalex.org/keywords/network-topology","display_name":"Network topology","score":0.43776246905326843},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.42827194929122925},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3724755644798279},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.34303218126296997},{"id":"https://openalex.org/keywords/topology","display_name":"Topology (electrical circuits)","score":0.34244590997695923},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.32793015241622925}],"concepts":[{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6357229948043823},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6284238696098328},{"id":"https://openalex.org/C68693459","wikidata":"https://www.wikidata.org/wiki/Q657586","display_name":"Hyperplane","level":2,"score":0.6068750619888306},{"id":"https://openalex.org/C186633575","wikidata":"https://www.wikidata.org/wiki/Q845060","display_name":"Maxima and minima","level":2,"score":0.5886716246604919},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.587266206741333},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.5627576112747192},{"id":"https://openalex.org/C21080849","wikidata":"https://www.wikidata.org/wiki/Q13611879","display_name":"Data point","level":2,"score":0.48823121190071106},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.44410115480422974},{"id":"https://openalex.org/C199845137","wikidata":"https://www.wikidata.org/wiki/Q145490","display_name":"Network topology","level":2,"score":0.43776246905326843},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.42827194929122925},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3724755644798279},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.34303218126296997},{"id":"https://openalex.org/C184720557","wikidata":"https://www.wikidata.org/wiki/Q7825049","display_name":"Topology (electrical circuits)","level":2,"score":0.34244590997695923},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.32793015241622925},{"id":"https://openalex.org/C111472728","wikidata":"https://www.wikidata.org/wiki/Q9471","display_name":"Epistemology","level":1,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0},{"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/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.3390/make4040051","is_oa":true,"landing_page_url":"https://doi.org/10.3390/make4040051","pdf_url":"https://www.mdpi.com/2504-4990/4/4/51/pdf?version=1668501014","source":{"id":"https://openalex.org/S4210213891","display_name":"Machine Learning and Knowledge Extraction","issn_l":"2504-4990","issn":["2504-4990"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310310987","host_organization_name":"Multidisciplinary Digital Publishing Institute","host_organization_lineage":["https://openalex.org/P4310310987"],"host_organization_lineage_names":["Multidisciplinary Digital Publishing Institute"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Machine Learning and Knowledge Extraction","raw_type":"journal-article"},{"id":"pmh:oai:doaj.org/article:ab08021471694729817e76407cba5709","is_oa":true,"landing_page_url":"https://doaj.org/article/ab08021471694729817e76407cba5709","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","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":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Machine Learning and Knowledge Extraction, Vol 4, Iss 4, Pp 1011-1023 (2022)","raw_type":"article"},{"id":"pmh:oai:mdpi.com:/2504-4990/4/4/51/","is_oa":true,"landing_page_url":"https://dx.doi.org/10.3390/make4040051","pdf_url":null,"source":{"id":"https://openalex.org/S4306400947","display_name":"MDPI (MDPI AG)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I4210097602","host_organization_name":"Multidisciplinary Digital Publishing Institute (Switzerland)","host_organization_lineage":["https://openalex.org/I4210097602"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Machine Learning and Knowledge Extraction; Volume 4; Issue 4; Pages: 1011-1023","raw_type":"Text"}],"best_oa_location":{"id":"doi:10.3390/make4040051","is_oa":true,"landing_page_url":"https://doi.org/10.3390/make4040051","pdf_url":"https://www.mdpi.com/2504-4990/4/4/51/pdf?version=1668501014","source":{"id":"https://openalex.org/S4210213891","display_name":"Machine Learning and Knowledge Extraction","issn_l":"2504-4990","issn":["2504-4990"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310310987","host_organization_name":"Multidisciplinary Digital Publishing Institute","host_organization_lineage":["https://openalex.org/P4310310987"],"host_organization_lineage_names":["Multidisciplinary Digital Publishing Institute"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Machine Learning and Knowledge Extraction","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4309315585.pdf","grobid_xml":"https://content.openalex.org/works/W4309315585.grobid-xml"},"referenced_works_count":22,"referenced_works":["https://openalex.org/W197865394","https://openalex.org/W1533861849","https://openalex.org/W1677182931","https://openalex.org/W1811750039","https://openalex.org/W1836465849","https://openalex.org/W1903029394","https://openalex.org/W2016043834","https://openalex.org/W2095705004","https://openalex.org/W2145173023","https://openalex.org/W2183341477","https://openalex.org/W2618169590","https://openalex.org/W2619516334","https://openalex.org/W2963351448","https://openalex.org/W2963446712","https://openalex.org/W2964054038","https://openalex.org/W2964292098","https://openalex.org/W2985382943","https://openalex.org/W3209948030","https://openalex.org/W4214745634","https://openalex.org/W4388297464","https://openalex.org/W6674330103","https://openalex.org/W6743923790"],"related_works":["https://openalex.org/W2375684291","https://openalex.org/W2369655046","https://openalex.org/W2019467317","https://openalex.org/W2354676191","https://openalex.org/W3188646203","https://openalex.org/W1585519779","https://openalex.org/W2909957174","https://openalex.org/W4246278799","https://openalex.org/W2141036804","https://openalex.org/W2944968625"],"abstract_inverted_index":{"Numerous":[0],"approaches":[1,256],"address":[2],"over-fitting":[3],"in":[4,137,193,235,259,275],"neural":[5],"networks:":[6],"by":[7,20,30,129,244],"imposing":[8],"a":[9,45,80,85,151,165,228,236,246],"penalty":[10],"on":[11,114,154],"the":[12,15,22,32,55,59,64,76,99,103,109,115,142,168,177,204,222,269],"parameters":[13,57],"of":[14,48,75,88,111,147,157,179,203,248],"network":[16,23,223,247],"(L1,":[17],"L2,":[18],"etc.);":[19,28],"changing":[21],"stochastically":[24],"(drop-out,":[25],"Gaussian":[26],"noise,":[27],"or":[29],"transforming":[31],"input":[33,140,159],"data":[34,89,160],"(batch":[35],"normalization,":[36],"etc.).":[37],"In":[38],"contrast,":[39],"we":[40,126,175],"aim":[41],"to":[42,58,70,98,121,133,164,215,278],"ensure":[43],"that":[44,72,92,128,185,209],"minimum":[46,86],"amount":[47],"supporting":[49],"evidence":[50],"is":[51,68,162,208,218,225],"present":[52],"when":[53,195,221],"fitting":[54,231],"model":[56],"training":[60,194,245,262],"data.":[61],"This,":[62],"at":[63,135],"single":[65],"neuron":[66,132,206],"level,":[67],"equivalent":[69,163],"ensuring":[71],"both":[73],"sides":[74],"separating":[77],"hyperplane":[78,152],"(for":[79],"standard":[81,116],"artificial":[82],"neuron)":[83],"have":[84],"number":[87],"points,":[90],"noting":[91],"these":[93],"points":[94],"need":[95],"not":[96,226],"belong":[97],"same":[100],"class":[101],"for":[102],"inner":[104],"layers.":[105],"We":[106,240],"firstly":[107],"benchmark":[108],"results":[110],"this":[112,173,197,242],"approach":[113],"Fashion-MINST":[117],"dataset,":[118],"comparing":[119],"it":[120],"various":[122],"regularization":[123,198,255,272],"techniques.":[124],"Interestingly,":[125],"note":[127],"nudging":[130],"each":[131,148],"divide,":[134],"least":[136],"part,":[138],"its":[139,158,276],"data,":[141],"resulting":[143],"networks":[144,211],"make":[145],"use":[146],"neuron,":[149],"avoiding":[150],"completely":[153],"one":[155],"side":[156],"(which":[161],"constant":[166,252],"into":[167],"next":[169],"layers).":[170],"To":[171],"illustrate":[172],"point,":[174],"study":[176],"prevalence":[178],"saturated":[180],"nodes":[181],"throughout":[182],"training,":[183],"showing":[184],"neurons":[186],"are":[187,212],"activated":[188],"more":[189],"frequently":[190],"and":[191,230],"earlier":[192],"using":[196],"approach.":[199],"A":[200],"direct":[201],"consequence":[202],"improved":[205],"activation":[207],"deep":[210,280],"now":[213],"easier":[214],"train.":[216],"This":[217],"crucially":[219],"important":[220],"topology":[224],"known":[227],"priori":[229],"often":[232],"remains":[233],"stuck":[234],"suboptimal":[237],"local":[238],"minima.":[239],"demonstrate":[241],"property":[243],"increasing":[249],"depth":[250],"(and":[251],"width);":[253],"most":[254],"will":[257],"result":[258],"increasingly":[260],"frequent":[261],"failures":[263],"(over":[264],"different":[265],"random":[266],"seeds),":[267],"whilst":[268],"proposed":[270],"evidence-based":[271],"significantly":[273],"outperforms":[274],"ability":[277],"train":[279],"networks.":[281]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":1}],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2022-11-26T00:00:00"}