{"id":"https://openalex.org/W3167758927","doi":"https://doi.org/10.1162/neco_a_01407","title":"Power Function Error Initialization Can Improve Convergence of Backpropagation Learning in Neural Networks for Classification","display_name":"Power Function Error Initialization Can Improve Convergence of Backpropagation Learning in Neural Networks for Classification","publication_year":2021,"publication_date":"2021-06-07","ids":{"openalex":"https://openalex.org/W3167758927","doi":"https://doi.org/10.1162/neco_a_01407","mag":"3167758927","pmid":"https://pubmed.ncbi.nlm.nih.gov/34310673"},"language":"en","primary_location":{"is_oa":false,"landing_page_url":"https://doi.org/10.1162/neco_a_01407","pdf_url":null,"source":{"id":"https://openalex.org/S207023548","display_name":"Neural Computation","issn_l":"0899-7667","issn":["0899-7667","1530-888X"],"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},"type":"article","type_crossref":"journal-article","indexed_in":["crossref","pubmed"],"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/A5032109856","display_name":"Andreas Knoblauch","orcid":"https://orcid.org/0000-0002-2534-0250"},"institutions":[{"id":"https://openalex.org/I4210124927","display_name":"Albstadt-Sigmaringen University","ror":"https://ror.org/03crxcn36","country_code":"DE","type":"education","lineage":["https://openalex.org/I4210124927"]}],"countries":["DE"],"is_corresponding":true,"raw_author_name":"Andreas Knoblauch","raw_affiliation_strings":["Albstadt-Sigmaringen University, Albstadt 72458, Germany knoblauch@hs-albsig.de"],"affiliations":[{"raw_affiliation_string":"Albstadt-Sigmaringen University, Albstadt 72458, Germany knoblauch@hs-albsig.de","institution_ids":["https://openalex.org/I4210124927"]}]}],"institution_assertions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5032109856"],"corresponding_institution_ids":["https://openalex.org/I4210124927"],"apc_list":null,"apc_paid":null,"fwci":0.685,"has_fulltext":false,"cited_by_count":6,"citation_normalized_percentile":{"value":0.793627,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":81,"max":83},"biblio":{"volume":"33","issue":"8","first_page":"2193","last_page":"2225"},"is_retracted":false,"is_paratext":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.9999,"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/T10320","display_name":"Neural Networks and Applications","score":0.9999,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9987,"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/T12676","display_name":"Machine Learning and ELM","score":0.9984,"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/initialization","display_name":"Initialization","score":0.9217421},{"id":"https://openalex.org/keywords/backpropagation","display_name":"Backpropagation","score":0.8766165},{"id":"https://openalex.org/keywords/error-function","display_name":"Error function","score":0.61508805},{"id":"https://openalex.org/keywords/cross-entropy","display_name":"Cross entropy","score":0.55087966},{"id":"https://openalex.org/keywords/activation-function","display_name":"Activation function","score":0.5019524},{"id":"https://openalex.org/keywords/supervised-learning","display_name":"Supervised Learning","score":0.43518385},{"id":"https://openalex.org/keywords/learning-rule","display_name":"Learning rule","score":0.4341054}],"concepts":[{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.9217421},{"id":"https://openalex.org/C155032097","wikidata":"https://www.wikidata.org/wiki/Q798503","display_name":"Backpropagation","level":3,"score":0.8766165},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.74054503},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6253116},{"id":"https://openalex.org/C202286095","wikidata":"https://www.wikidata.org/wiki/Q579262","display_name":"Error function","level":2,"score":0.61508805},{"id":"https://openalex.org/C167981619","wikidata":"https://www.wikidata.org/wiki/Q1685498","display_name":"Cross entropy","level":3,"score":0.55087966},{"id":"https://openalex.org/C106301342","wikidata":"https://www.wikidata.org/wiki/Q4117933","display_name":"Entropy (arrow of time)","level":2,"score":0.5102494},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.50364894},{"id":"https://openalex.org/C38365724","wikidata":"https://www.wikidata.org/wiki/Q4677469","display_name":"Activation function","level":3,"score":0.5019524},{"id":"https://openalex.org/C139945424","wikidata":"https://www.wikidata.org/wiki/Q1940696","display_name":"Mean squared error","level":2,"score":0.49796534},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.4855199},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.43518385},{"id":"https://openalex.org/C2779127903","wikidata":"https://www.wikidata.org/wiki/Q6510194","display_name":"Learning rule","level":3,"score":0.4341054},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.41437757},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.38547182},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.32587096},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.30614585},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.113257945},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"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/C78458016","wikidata":"https://www.wikidata.org/wiki/Q840400","display_name":"Evolutionary biology","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","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}],"mesh":[{"descriptor_ui":"D000465","descriptor_name":"Algorithms","qualifier_ui":"","qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D016571","descriptor_name":"Neural Networks, Computer","qualifier_ui":"","qualifier_name":null,"is_major_topic":true},{"descriptor_ui":"D011336","descriptor_name":"Probability","qualifier_ui":"","qualifier_name":null,"is_major_topic":false}],"locations_count":2,"locations":[{"is_oa":false,"landing_page_url":"https://doi.org/10.1162/neco_a_01407","pdf_url":null,"source":{"id":"https://openalex.org/S207023548","display_name":"Neural Computation","issn_l":"0899-7667","issn":["0899-7667","1530-888X"],"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},{"is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/34310673","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":["National Institutes of Health"],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false}],"best_oa_location":null,"sustainable_development_goals":[],"grants":[],"datasets":[],"versions":[],"referenced_works_count":49,"referenced_works":["https://openalex.org/W1474440726","https://openalex.org/W1482764590","https://openalex.org/W1522301498","https://openalex.org/W1533861849","https://openalex.org/W1573503290","https://openalex.org/W1601795611","https://openalex.org/W1652505363","https://openalex.org/W1677182931","https://openalex.org/W1689711448","https://openalex.org/W1726370773","https://openalex.org/W1836465849","https://openalex.org/W1971129545","https://openalex.org/W1981577790","https://openalex.org/W1984741961","https://openalex.org/W1990147185","https://openalex.org/W1995539593","https://openalex.org/W1999001021","https://openalex.org/W2006275398","https://openalex.org/W2011815965","https://openalex.org/W2018435387","https://openalex.org/W2032498155","https://openalex.org/W2064675550","https://openalex.org/W2076063813","https://openalex.org/W2095425517","https://openalex.org/W2099111195","https://openalex.org/W2107878631","https://openalex.org/W2112796928","https://openalex.org/W2126076951","https://openalex.org/W2140833774","https://openalex.org/W2151718948","https://openalex.org/W2194775991","https://openalex.org/W2288231337","https://openalex.org/W2523246573","https://openalex.org/W2559655401","https://openalex.org/W2578069546","https://openalex.org/W2594265094","https://openalex.org/W2899790086","https://openalex.org/W2919115771","https://openalex.org/W2953384591","https://openalex.org/W2964121744","https://openalex.org/W2964309882","https://openalex.org/W2984522809","https://openalex.org/W2993383518","https://openalex.org/W3118608800","https://openalex.org/W3122520870","https://openalex.org/W4211143448","https://openalex.org/W4221067452","https://openalex.org/W4297748376","https://openalex.org/W603115923"],"related_works":["https://openalex.org/W3209256591","https://openalex.org/W3175626380","https://openalex.org/W3167758927","https://openalex.org/W2895200962","https://openalex.org/W2609564064","https://openalex.org/W2391384657","https://openalex.org/W2368205053","https://openalex.org/W2244330272","https://openalex.org/W2145489748","https://openalex.org/W2139855773"],"abstract_inverted_index":{"Supervised":[0],"learning":[1,111],"corresponds":[2],"to":[3,97,150,159],"minimizing":[4],"a":[5,83,98,115,186,191],"loss":[6,71,102,154,201],"or":[7,75],"cost":[8],"function":[9],"expressing":[10],"the":[11,18,24,36,41,45,53,60,123,145,152,160,166,172,178],"differences":[12],"between":[13],"model":[14],"predictions":[15],"yn":[16],"and":[17,125,134,169],"target":[19],"values":[20],"tn":[21],"given":[22],"by":[23,59],"training":[25],"data.":[26],"In":[27],"neural":[28,136,195],"networks,":[29],"this":[30],"means":[31],"backpropagating":[32],"error":[33,50,86,163,180],"signals":[34,51,164],"through":[35],"transposed":[37],"weight":[38],"matrixes":[39],"from":[40],"output":[42,54,167,196],"layer":[43,55,168],"toward":[44],"input":[46],"layer.":[47],"For":[48],"this,":[49],"in":[52,130,132,165],"are":[56],"typically":[57],"initialized":[58],"difference":[61],"yn-":[62],"tn,":[63],"which":[64],"is":[65],"optimal":[66],"for":[67,94,144],"several":[68],"commonly":[69],"used":[70],"functions":[72,91,103,155],"like":[73],"cross-entropy":[74],"sum":[76],"of":[77,101,118,127,162,194],"squared":[78],"errors.":[79],"Here":[80],"I":[81],"evaluate":[82],"more":[84,175],"general":[85],"initialization":[87,181],"method":[88],"using":[89],"power":[90],"|yn-":[92],"tn|q":[93],"q>0,":[95],"corresponding":[96],"new":[99,153,179],"family":[100],"that":[104,114],"generalize":[105],"cross-entropy.":[106],"Surprisingly,":[107],"experiments":[108],"on":[109],"various":[110],"tasks":[112],"reveal":[113],"proper":[116],"choice":[117],"q":[119],"can":[120],"significantly":[121],"improve":[122],"speed":[124],"convergence":[126],"backpropagation":[128],"learning,":[129],"particular":[131],"deep":[133],"recurrent":[135],"networks.":[137],"The":[138],"results":[139],"suggest":[140],"two":[141],"main":[142],"reasons":[143],"observed":[146],"improvements.":[147],"First,":[148],"compared":[149],"cross-entropy,":[151],"provide":[156,185],"better":[157,187],"fits":[158],"distribution":[161],"therefore":[170],"maximize":[171],"model's":[173],"likelihood":[174],"efficiently.":[176],"Second,":[177],"procedure":[182],"may":[183],"often":[184],"gradient-to-loss":[188],"ratio":[189],"over":[190],"broad":[192],"range":[193],"activity,":[197],"thereby":[198],"avoiding":[199],"flat":[200],"landscapes":[202],"with":[203],"vanishing":[204],"gradients.":[205]},"cited_by_api_url":"https://api.openalex.org/works?filter=cites:W3167758927","counts_by_year":[{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":4}],"updated_date":"2025-01-09T02:02:28.333844","created_date":"2021-06-22"}