{"id":"https://openalex.org/W4323240661","doi":"https://doi.org/10.5220/0011678000003411","title":"Gradient Clipping in Deep Learning: A Dynamical Systems Perspective","display_name":"Gradient Clipping in Deep Learning: A Dynamical Systems Perspective","publication_year":2023,"publication_date":"2023-01-01","ids":{"openalex":"https://openalex.org/W4323240661","doi":"https://doi.org/10.5220/0011678000003411"},"language":"en","primary_location":{"id":"doi:10.5220/0011678000003411","is_oa":true,"landing_page_url":"https://doi.org/10.5220/0011678000003411","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.5220/0011678000003411","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5086784011","display_name":"Arunselvan Ramaswamy","orcid":"https://orcid.org/0000-0001-7547-8111"},"institutions":[{"id":"https://openalex.org/I43968019","display_name":"Karlstad University","ror":"https://ror.org/05s754026","country_code":"SE","type":"education","lineage":["https://openalex.org/I43968019"]}],"countries":["SE"],"is_corresponding":true,"raw_author_name":"Arunselvan Ramaswamy","raw_affiliation_strings":["Dept. of Mathematics and Computer Science, Karlstad University, 651 88 Karlstad, Sweden, --- Select a Country ---"],"affiliations":[{"raw_affiliation_string":"Dept. of Mathematics and Computer Science, Karlstad University, 651 88 Karlstad, Sweden, --- Select a Country ---","institution_ids":["https://openalex.org/I43968019"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5086784011"],"corresponding_institution_ids":["https://openalex.org/I43968019"],"apc_list":null,"apc_paid":null,"fwci":1.0784,"has_fulltext":true,"cited_by_count":6,"citation_normalized_percentile":{"value":0.80666045,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":97,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"107","last_page":"114"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.6432999968528748,"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":0.6432999968528748,"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/clipping","display_name":"Clipping (morphology)","score":0.6823011040687561},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.6787776947021484},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5437440276145935},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.46153247356414795},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4231386184692383},{"id":"https://openalex.org/keywords/philosophy","display_name":"Philosophy","score":0.10034865140914917}],"concepts":[{"id":"https://openalex.org/C2776848632","wikidata":"https://www.wikidata.org/wiki/Q853463","display_name":"Clipping (morphology)","level":2,"score":0.6823011040687561},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.6787776947021484},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5437440276145935},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.46153247356414795},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4231386184692383},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.10034865140914917},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5220/0011678000003411","is_oa":true,"landing_page_url":"https://doi.org/10.5220/0011678000003411","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods","raw_type":"proceedings-article"},{"id":"pmh:oai:DiVA.org:kau-97507","is_oa":true,"landing_page_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kau:diva-97507","pdf_url":"https://kau.diva-portal.org/smash/get/diva2:1814997/FULLTEXT01","source":{"id":"https://openalex.org/S4306402134","display_name":"Digitala vetenskapliga arkivet (Diva) (Karlstad University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I43968019","host_organization_name":"Karlstad University","host_organization_lineage":["https://openalex.org/I43968019"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"doi:10.5220/0011678000003411","is_oa":true,"landing_page_url":"https://doi.org/10.5220/0011678000003411","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":15,"referenced_works":["https://openalex.org/W1514218977","https://openalex.org/W1580445772","https://openalex.org/W1601795611","https://openalex.org/W1964757264","https://openalex.org/W2076063813","https://openalex.org/W2293415235","https://openalex.org/W2576544140","https://openalex.org/W2595897327","https://openalex.org/W2962716894","https://openalex.org/W2962741973","https://openalex.org/W3005560345","https://openalex.org/W3153977456","https://openalex.org/W4285819331","https://openalex.org/W4289420852","https://openalex.org/W4391410851"],"related_works":["https://openalex.org/W2731899572","https://openalex.org/W3215138031","https://openalex.org/W3009238340","https://openalex.org/W2939353110","https://openalex.org/W4321369474","https://openalex.org/W4360585206","https://openalex.org/W4285208911","https://openalex.org/W3082895349","https://openalex.org/W4213079790","https://openalex.org/W2248239756"],"abstract_inverted_index":{"Neural":[0],"networks":[1],"are":[2,45,62],"ubiquitous":[3],"components":[4],"of":[5,122,155,164],"Machine":[6],"Learning":[7],"(ML)":[8],"algorithms.":[9],"However,":[10],"training":[11,131,148],"them":[12],"is":[13,26,73,139,152],"challenging":[14],"due":[15],"to":[16,28,49,119,142],"problems":[17],"associated":[18],"with":[19,108,140],"exploding":[20,37],"and":[21,35,94],"vanishing":[22,33],"loss-gradients.":[23],"Gradient":[24],"clipping":[25,87,92,98,104,113],"shown":[27],"effectively":[29],"combat":[30],"both":[31],"the":[32,36,41,54,95,128,143,147,162,168],"gradients":[34,38,44],"problems.":[39],"As":[40],"name":[42],"suggests,":[43],"clipped":[46],"in":[47,70,133],"order":[48],"prevent":[50],"large":[51],"updates.":[52],"At":[53],"same":[55],"time,":[56],"very":[57,74,109],"small":[58],"neural":[59,123],"network":[60,124],"weights":[61,125],"updated":[63],"using":[64],"larger":[65],"step-sizes.":[66],"Although":[67],"widely":[68],"used":[69],"practice,":[71],"there":[72],"little":[75],"theory":[76,163],"surrounding":[77],"clipping.":[78,157],"In":[79],"this":[80],"paper,":[81],"we":[82],"analyze":[83],"two":[84],"popular":[85],"gradient":[86,91,97,103,116,156],"techniques-the":[88],"classic":[89],"norm-based":[90],"method":[93],"adaptive":[96],"technique.":[99],"We":[100,158],"prove":[101],"that":[102,126,145],"ensures":[105],"numerical":[106],"stability":[107],"high":[110],"probability.":[111],"Further,":[112],"based":[114],"stochastic":[115],"descent":[117],"converges":[118],"a":[120,134,153],"set":[121],"minimizes":[127],"average":[129],"scaled":[130],"loss":[132],"local":[135],"sense.":[136],"The":[137,150],"averaging":[138],"respect":[141],"distribution":[144],"generated":[146],"data.":[149],"scaling":[151],"consequence":[154],"use":[159],"tools":[160],"from":[161],"dynamical":[165],"systems":[166],"for":[167],"presented":[169],"analysis.":[170]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":5}],"updated_date":"2026-03-11T14:59:36.786465","created_date":"2025-10-10T00:00:00"}