{"id":"https://openalex.org/W2948746540","doi":null,"title":"Convex Optimization using Sparsified Stochastic Gradient Descent with Memory","display_name":"Convex Optimization using Sparsified Stochastic Gradient Descent with Memory","publication_year":2018,"publication_date":"2018-06-27","ids":{"openalex":"https://openalex.org/W2948746540","mag":"2948746540"},"language":"en","primary_location":{"id":"pmh:oai:infoscience.epfl.ch:262760","doi":null,"is_oa":true,"landing_page_url":"https://infoscience.epfl.ch/handle/20.500.14299/153353","pdf_url":"https://infoscience.epfl.ch/record/262760/files/master-thesis.pdf","source":{"id":"https://openalex.org/S4306400487","display_name":"Infoscience (Ecole Polytechnique F\u00e9d\u00e9rale de Lausanne)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I5124864","host_organization_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","host_organization_lineage":["https://openalex.org/I5124864"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"master thesis"},"volume":null,"type":"dissertation","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://infoscience.epfl.ch/record/262760/files/master-thesis.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5080236521","display_name":"Jean-Baptiste Cordonnier","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Jean-Baptiste Cordonnier","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5080236521"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":5,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9994999766349792,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9994999766349792,"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/T10249","display_name":"Distributed Control Multi-Agent Systems","score":9.999999747378752e-05,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10963","display_name":"Advanced Optimization Algorithms Research","score":0.0,"subfield":{"id":"https://openalex.org/subfields/2612","display_name":"Numerical Analysis"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/distributed-stochastic-optimization","display_name":"distributed stochastic optimization","score":0.7429999709129333},{"id":"https://openalex.org/keywords/communication-efficiency","display_name":"communication efficiency","score":0.7049999833106995},{"id":"https://openalex.org/keywords/top-k-sparsification","display_name":"top-k sparsification","score":0.7789999842643738},{"id":"https://openalex.org/keywords/convergence-rate","display_name":"convergence rate","score":0.6679999828338623},{"id":"https://openalex.org/keywords/gradient-quantization","display_name":"gradient quantization","score":0.492000013589859},{"id":"https://openalex.org/keywords/convex-optimization","display_name":"convex optimization","score":0.4970000088214874}],"concepts":[{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.7131390571594238},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.49773815274238586},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4745557904243469},{"id":"https://openalex.org/C112680207","wikidata":"https://www.wikidata.org/wiki/Q714886","display_name":"Regular polygon","level":2,"score":0.45828479528427124},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.4510178565979004},{"id":"https://openalex.org/C10494615","wikidata":"https://www.wikidata.org/wiki/Q17086765","display_name":"Proximal Gradient Methods","level":4,"score":0.4505614638328552},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4073023200035095},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.3593711853027344},{"id":"https://openalex.org/C157972887","wikidata":"https://www.wikidata.org/wiki/Q463359","display_name":"Convex optimization","level":3,"score":0.3440467119216919},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.27980196475982666},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.11311841011047363},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:infoscience.epfl.ch:262760","doi":null,"is_oa":true,"landing_page_url":"https://infoscience.epfl.ch/handle/20.500.14299/153353","pdf_url":"https://infoscience.epfl.ch/record/262760/files/master-thesis.pdf","source":{"id":"https://openalex.org/S4306400487","display_name":"Infoscience (Ecole Polytechnique F\u00e9d\u00e9rale de Lausanne)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I5124864","host_organization_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","host_organization_lineage":["https://openalex.org/I5124864"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"master thesis"},{"id":"mag:2948746540","doi":null,"is_oa":false,"landing_page_url":"https://infoscience.epfl.ch/record/262760","pdf_url":null,"source":null,"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":null}],"best_oa_location":{"id":"pmh:oai:infoscience.epfl.ch:262760","doi":null,"is_oa":true,"landing_page_url":"https://infoscience.epfl.ch/handle/20.500.14299/153353","pdf_url":"https://infoscience.epfl.ch/record/262760/files/master-thesis.pdf","source":{"id":"https://openalex.org/S4306400487","display_name":"Infoscience (Ecole Polytechnique F\u00e9d\u00e9rale de Lausanne)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I5124864","host_organization_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","host_organization_lineage":["https://openalex.org/I5124864"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"master thesis"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2948746540.pdf","grobid_xml":"https://content.openalex.org/works/W2948746540.grobid-xml"},"referenced_works_count":7,"referenced_works":["https://openalex.org/W1757796397","https://openalex.org/W1791038712","https://openalex.org/W1843828865","https://openalex.org/W2460815216","https://openalex.org/W2622263826","https://openalex.org/W2963408258","https://openalex.org/W6891860974"],"related_works":["https://openalex.org/W2890924858","https://openalex.org/W2766140019","https://openalex.org/W2981202102","https://openalex.org/W2891970177","https://openalex.org/W2786546591","https://openalex.org/W2948092338","https://openalex.org/W2966527647","https://openalex.org/W6929402797","https://openalex.org/W3092862163","https://openalex.org/W2891447341"],"abstract_inverted_index":{"The":[0,123],"interest":[1],"for":[2,108,180],"distributed":[3,18,42],"stochastic":[4],"optimization":[5],"has":[6],"raised":[7],"to":[8,47,76,127,148,201],"train":[9],"complex":[10],"Machine":[11],"Learning":[12],"models":[13],"with":[14,165,168],"more":[15],"data":[16,126],"on":[17,212],"systems.":[19],"Increasing":[20],"the":[21,26,38,72,88,134,158,176,199,216,219],"computation":[22],"power":[23],"speeds":[24],"up":[25,156],"training":[27],"but":[28],"it":[29,96],"faces":[30],"a":[31,98,105,112,153],"communication":[32],"bottleneck":[33],"between":[34],"workers":[35],"which":[36],"hurts":[37],"scale-up":[39],"of":[40,116,125,139,160,193,218],"these":[41],"algorithms.":[43],"Previous":[44],"work":[45],"tried":[46],"address":[48],"this":[49],"issue":[50],"through":[51,67],"quantization":[52],"by":[53,69],"broadcasting":[54],"low":[55],"precision":[56],"updates":[57,221],"(Seide":[58],"et":[59,63,83,209],"al.,":[60,64,84],"2014;":[61],"Alistarh":[62],"2017)":[65],"or":[66],"sparsification":[68,106],"sharing":[70],"only":[71,111],"most":[73],"important":[74],"coordinates":[75,117],"update":[77],"(Aji":[78],"and":[79,151,172,174,182],"Heafield,":[80],"2017;":[81],"Lin":[82],"2018).":[85],"Even":[86],"though":[87],"sparsifica-":[89],"tion":[90],"method":[91],"works":[92],"well":[93],"in":[94,157,170,189],"practice,":[95],"lacked":[97],"theoretical":[99],"proof":[100,146],"until":[101],"now.":[102],"We":[103,162],"propose":[104],"scheme":[107,186],"SGD":[109,141,167],"where":[110],"small":[113],"constant":[114],"number":[115,159,192],"are":[118],"applied":[119],"at":[120],"each":[121],"iteration.":[122],"amount":[124],"be":[128],"communicated":[129],"is":[130,142,222],"drastically":[131],"reduced":[132],"while":[133],"O(1/T":[135],")":[136],"convergence":[137,178],"rate":[138],"vanilla":[140],"preserved.":[143],"Our":[144,185],"concise":[145],"extends":[147],"parallel":[149],"setting":[150],"gains":[152],"linear":[154],"speed":[155],"workers.":[161],"exper-":[163],"iment":[164],"sparsified":[166],"memory":[169],"C++":[171],"Python,":[173],"show":[175],"excellent":[177],"properties":[179],"top-k":[181],"random-k":[183],"operators.":[184],"outperforms":[187],"QSGD":[188],"progress":[190],"per":[191],"bits":[194],"sent.":[195],"It":[196],"also":[197],"opens":[198],"path":[200],"using":[202],"lock-free":[203],"asynchronous":[204],"parallelization":[205],"(e.g.":[206],"Hogwild!":[207],"Niu":[208],"al.":[210],"(2011))":[211],"dense":[213],"problems":[214],"as":[215],"sparsity":[217],"gradient":[220],"enforced.":[223]},"counts_by_year":[{"year":2020,"cited_by_count":3},{"year":2019,"cited_by_count":2}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2019-06-14T00:00:00"}