{"id":"https://openalex.org/W4206722564","doi":"https://doi.org/10.23919/eusipco54536.2021.9616015","title":"Classification Error Approximation of a Compressed Linear Softmax Layer","display_name":"Classification Error Approximation of a Compressed Linear Softmax Layer","publication_year":2021,"publication_date":"2021-08-23","ids":{"openalex":"https://openalex.org/W4206722564","doi":"https://doi.org/10.23919/eusipco54536.2021.9616015"},"language":"en","primary_location":{"id":"doi:10.23919/eusipco54536.2021.9616015","is_oa":false,"landing_page_url":"https://doi.org/10.23919/eusipco54536.2021.9616015","pdf_url":null,"source":{"id":"https://openalex.org/S4363607854","display_name":"2021 29th European Signal Processing Conference (EUSIPCO)","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":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 29th European Signal Processing Conference (EUSIPCO)","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/A5016595090","display_name":"Diana Resmerita","orcid":null},"institutions":[{"id":"https://openalex.org/I1294671590","display_name":"Centre National de la Recherche Scientifique","ror":"https://ror.org/02feahw73","country_code":"FR","type":"funder","lineage":["https://openalex.org/I1294671590"]},{"id":"https://openalex.org/I201841394","display_name":"Universit\u00e9 C\u00f4te d'Azur","ror":"https://ror.org/019tgvf94","country_code":"FR","type":"education","lineage":["https://openalex.org/I201841394"]}],"countries":["FR"],"is_corresponding":true,"raw_author_name":"Diana Resmerita","raw_affiliation_strings":["Universit\u00e9 C\u00f4te d'Azur, CNRS, I3S and Kalray, France"],"affiliations":[{"raw_affiliation_string":"Universit\u00e9 C\u00f4te d'Azur, CNRS, I3S and Kalray, France","institution_ids":["https://openalex.org/I201841394","https://openalex.org/I1294671590"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043594332","display_name":"Rodrigo Cabral Farias","orcid":"https://orcid.org/0000-0001-9261-6037"},"institutions":[{"id":"https://openalex.org/I201841394","display_name":"Universit\u00e9 C\u00f4te d'Azur","ror":"https://ror.org/019tgvf94","country_code":"FR","type":"education","lineage":["https://openalex.org/I201841394"]},{"id":"https://openalex.org/I1294671590","display_name":"Centre National de la Recherche Scientifique","ror":"https://ror.org/02feahw73","country_code":"FR","type":"funder","lineage":["https://openalex.org/I1294671590"]}],"countries":["FR"],"is_corresponding":false,"raw_author_name":"Rodrigo Cabral Farias","raw_affiliation_strings":["Universit\u00e9 C\u00f4te d'Azur, CNRS, I3S and Kalray, France"],"affiliations":[{"raw_affiliation_string":"Universit\u00e9 C\u00f4te d'Azur, CNRS, I3S and Kalray, France","institution_ids":["https://openalex.org/I201841394","https://openalex.org/I1294671590"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5045583054","display_name":"Lionel Fillatre","orcid":"https://orcid.org/0000-0002-8152-1769"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lionel Fillatre","raw_affiliation_strings":["Beno\u00eet Dupont de Dinechin Kalray,France"],"affiliations":[{"raw_affiliation_string":"Beno\u00eet Dupont de Dinechin Kalray,France","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5016595090"],"corresponding_institution_ids":["https://openalex.org/I1294671590","https://openalex.org/I201841394"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.20583287,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1446","last_page":"1450"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9990000128746033,"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/T11689","display_name":"Adversarial Robustness in Machine 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/T10036","display_name":"Advanced Neural Network Applications","score":0.9980000257492065,"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/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.9977999925613403,"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/softmax-function","display_name":"Softmax function","score":0.873570442199707},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6631354689598083},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.5917605757713318},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5274149179458618},{"id":"https://openalex.org/keywords/distortion","display_name":"Distortion (music)","score":0.50301593542099},{"id":"https://openalex.org/keywords/multivariate-statistics","display_name":"Multivariate statistics","score":0.4731296896934509},{"id":"https://openalex.org/keywords/approximation-algorithm","display_name":"Approximation algorithm","score":0.45013490319252014},{"id":"https://openalex.org/keywords/computational-complexity-theory","display_name":"Computational complexity theory","score":0.44832390546798706},{"id":"https://openalex.org/keywords/heuristic","display_name":"Heuristic","score":0.4378044903278351},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4182452857494354},{"id":"https://openalex.org/keywords/data-compression","display_name":"Data compression","score":0.41693875193595886},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.41406241059303284},{"id":"https://openalex.org/keywords/linear-approximation","display_name":"Linear approximation","score":0.41086894273757935},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3489072322845459},{"id":"https://openalex.org/keywords/bandwidth","display_name":"Bandwidth (computing)","score":0.2305225133895874},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.22328972816467285},{"id":"https://openalex.org/keywords/nonlinear-system","display_name":"Nonlinear system","score":0.08606523275375366},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.08328935503959656}],"concepts":[{"id":"https://openalex.org/C188441871","wikidata":"https://www.wikidata.org/wiki/Q7554146","display_name":"Softmax function","level":3,"score":0.873570442199707},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6631354689598083},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5917605757713318},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5274149179458618},{"id":"https://openalex.org/C126780896","wikidata":"https://www.wikidata.org/wiki/Q899871","display_name":"Distortion (music)","level":4,"score":0.50301593542099},{"id":"https://openalex.org/C161584116","wikidata":"https://www.wikidata.org/wiki/Q1952580","display_name":"Multivariate statistics","level":2,"score":0.4731296896934509},{"id":"https://openalex.org/C148764684","wikidata":"https://www.wikidata.org/wiki/Q621751","display_name":"Approximation algorithm","level":2,"score":0.45013490319252014},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.44832390546798706},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.4378044903278351},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4182452857494354},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.41693875193595886},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41406241059303284},{"id":"https://openalex.org/C160824197","wikidata":"https://www.wikidata.org/wiki/Q2071054","display_name":"Linear approximation","level":3,"score":0.41086894273757935},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3489072322845459},{"id":"https://openalex.org/C2776257435","wikidata":"https://www.wikidata.org/wiki/Q1576430","display_name":"Bandwidth (computing)","level":2,"score":0.2305225133895874},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.22328972816467285},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.08606523275375366},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.08328935503959656},{"id":"https://openalex.org/C194257627","wikidata":"https://www.wikidata.org/wiki/Q211554","display_name":"Amplifier","level":3,"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/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.23919/eusipco54536.2021.9616015","is_oa":false,"landing_page_url":"https://doi.org/10.23919/eusipco54536.2021.9616015","pdf_url":null,"source":{"id":"https://openalex.org/S4363607854","display_name":"2021 29th European Signal Processing Conference (EUSIPCO)","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":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 29th European Signal Processing Conference (EUSIPCO)","raw_type":"proceedings-article"},{"id":"pmh:oai:HAL:hal-03624036v1","is_oa":false,"landing_page_url":"https://hal.science/hal-03624036","pdf_url":null,"source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"2021 29th European Signal Processing Conference (EUSIPCO), Aug 2021, Dublin, Ireland. pp.1446-1450, ⟨10.23919/EUSIPCO54536.2021.9616015⟩","raw_type":"Conference papers"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":35,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W1724438581","https://openalex.org/W1976990135","https://openalex.org/W2018345399","https://openalex.org/W2167901303","https://openalex.org/W2194775991","https://openalex.org/W2276892413","https://openalex.org/W2279098554","https://openalex.org/W2300242332","https://openalex.org/W2325756088","https://openalex.org/W2392262332","https://openalex.org/W2478708596","https://openalex.org/W2619096655","https://openalex.org/W2787884921","https://openalex.org/W2809624076","https://openalex.org/W2896737121","https://openalex.org/W2963122961","https://openalex.org/W2963236897","https://openalex.org/W2982644126","https://openalex.org/W3105986941","https://openalex.org/W3133603318","https://openalex.org/W3193530571","https://openalex.org/W4246040193","https://openalex.org/W4295262505","https://openalex.org/W4297775537","https://openalex.org/W4299322810","https://openalex.org/W6637373629","https://openalex.org/W6637709462","https://openalex.org/W6693397755","https://openalex.org/W6698200048","https://openalex.org/W6737664043","https://openalex.org/W6738642365","https://openalex.org/W6748600614","https://openalex.org/W6753069482","https://openalex.org/W6755737014"],"related_works":["https://openalex.org/W2017216075","https://openalex.org/W3122413977","https://openalex.org/W96441109","https://openalex.org/W143765064","https://openalex.org/W4391069889","https://openalex.org/W2885465588","https://openalex.org/W189995875","https://openalex.org/W2963656294","https://openalex.org/W2163172376","https://openalex.org/W1974461541"],"abstract_inverted_index":{"Deep":[0],"neural":[1,98],"networks":[2],"need":[3],"to":[4,8,22],"be":[5,140],"compressed":[6,101],"due":[7],"their":[9],"high":[10],"memory":[11],"requirements":[12],"and":[13,73,125],"computational":[14],"complexity.":[15],"Previous":[16],"papers":[17],"have":[18],"proposed":[19,134],"numerous":[20],"methods":[21],"quantize":[23],"the":[24,33,39,52,63,66,75,90,103,106,118,136],"weights":[25],"with":[26],"a":[27,45,57,70,86,96,110],"single":[28],"bit":[29],"or":[30],"more.":[31],"However,":[32],"loss":[34],"of":[35,48,69,95],"accuracy":[36],"involved":[37],"in":[38,135],"compression":[40],"is":[41,80,100],"scarcely":[42],"studied":[43],"from":[44],"theoretical":[46,87],"point":[47],"view.":[49],"Motivated":[50],"by":[51],"rate-distortion":[53],"theory,":[54],"we":[55,84],"propose":[56],"new":[58],"distortion":[59,79],"measure":[60],"which":[61],"assesses":[62],"gap":[64],"between":[65],"Bayes":[67],"risk":[68],"classifier":[71],"before":[72],"after":[74],"compression.":[76],"Since":[77],"this":[78],"not":[81,139],"easily":[82],"tractable,":[83],"derive":[85],"approximation":[88,119],"when":[89],"last":[91],"fully":[92],"connected":[93],"layer":[94,107],"deep":[97],"network":[99],"under":[102],"assumption":[104],"that":[105,117,131],"inputs":[108],"follow":[109],"multivariate":[111],"normal":[112],"distribution.":[113],"Numerical":[114],"results":[115],"show":[116,130],"performs":[120],"well":[121],"on":[122],"both":[123],"synthetic":[124],"real":[126],"data.":[127],"We":[128],"also":[129],"heuristic":[132],"quantizers":[133],"literature":[137],"may":[138],"optimal.":[141]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}