{"id":"https://openalex.org/W4296415412","doi":"https://doi.org/10.48550/arxiv.2209.07670","title":"Reducing Variance in Temporal-Difference Value Estimation via Ensemble of Deep Networks","display_name":"Reducing Variance in Temporal-Difference Value Estimation via Ensemble of Deep Networks","publication_year":2022,"publication_date":"2022-09-16","ids":{"openalex":"https://openalex.org/W4296415412","doi":"https://doi.org/10.48550/arxiv.2209.07670"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2209.07670","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2209.07670","pdf_url":"https://arxiv.org/pdf/2209.07670","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"text"},"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2209.07670","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5008314736","display_name":"Litian Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Liang, Litian","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004875764","display_name":"Yaosheng Xu","orcid":"https://orcid.org/0000-0002-2187-7157"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Yaosheng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064779744","display_name":"Stephen McAleer","orcid":"https://orcid.org/0000-0003-0118-6874"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"McAleer, Stephen","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072406917","display_name":"Dailin Hu","orcid":"https://orcid.org/0000-0002-8462-1095"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Dailin","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048047705","display_name":"Alexander Ihler","orcid":"https://orcid.org/0000-0002-4331-1015"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ihler, Alexander","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049349154","display_name":"Pieter Abbeel","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Abbeel, Pieter","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5111668270","display_name":"Roy Fox","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fox, Roy","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5008314736"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":4,"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/T13283","display_name":"Mental Health Research Topics","score":0.9606000185012817,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T13283","display_name":"Mental Health Research Topics","score":0.9606000185012817,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.7420509457588196},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7221048474311829},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.6468724608421326},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.5021610260009766},{"id":"https://openalex.org/keywords/ensemble-learning","display_name":"Ensemble learning","score":0.49633556604385376},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.4815540611743927},{"id":"https://openalex.org/keywords/lagging","display_name":"Lagging","score":0.4659487009048462},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.46539318561553955},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.4525434970855713},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.4424664080142975},{"id":"https://openalex.org/keywords/sample-size-determination","display_name":"Sample size determination","score":0.4247822165489197},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.4186446964740753},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3294884264469147},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.19737768173217773}],"concepts":[{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.7420509457588196},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7221048474311829},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.6468724608421326},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.5021610260009766},{"id":"https://openalex.org/C45942800","wikidata":"https://www.wikidata.org/wiki/Q245652","display_name":"Ensemble learning","level":2,"score":0.49633556604385376},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.4815540611743927},{"id":"https://openalex.org/C2776962539","wikidata":"https://www.wikidata.org/wiki/Q6472078","display_name":"Lagging","level":2,"score":0.4659487009048462},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46539318561553955},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4525434970855713},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4424664080142975},{"id":"https://openalex.org/C129848803","wikidata":"https://www.wikidata.org/wiki/Q2564360","display_name":"Sample size determination","level":2,"score":0.4247822165489197},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.4186446964740753},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3294884264469147},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.19737768173217773},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C121955636","wikidata":"https://www.wikidata.org/wiki/Q4116214","display_name":"Accounting","level":1,"score":0.0},{"id":"https://openalex.org/C43617362","wikidata":"https://www.wikidata.org/wiki/Q170050","display_name":"Chromatography","level":1,"score":0.0},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.0},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.0},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2209.07670","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2209.07670","pdf_url":"https://arxiv.org/pdf/2209.07670","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"text"},{"id":"doi:10.48550/arxiv.2209.07670","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2209.07670","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article-journal"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2209.07670","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2209.07670","pdf_url":"https://arxiv.org/pdf/2209.07670","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2135562825","https://openalex.org/W3145783610","https://openalex.org/W2354181859","https://openalex.org/W2757096358","https://openalex.org/W2472614530","https://openalex.org/W2375646802","https://openalex.org/W2052668055","https://openalex.org/W2162038069","https://openalex.org/W627568556","https://openalex.org/W2221502901"],"abstract_inverted_index":{"In":[0,50],"temporal-difference":[1],"reinforcement":[2],"learning":[3,39],"algorithms,":[4],"variance":[5,43,95],"in":[6,37,75,122,154,170],"value":[7],"estimation":[8,42,94],"can":[9],"cause":[10,47],"instability":[11],"and":[12,109,117,157,173,178],"overestimation":[13],"of":[14,48,89,107,127,134],"the":[15,45,78,98,119,125,145],"maximal":[16],"target":[17,62,100],"value.":[18],"Many":[19],"algorithms":[20],"have":[21,34],"been":[22],"proposed":[23],"to":[24,96],"reduce":[25],"overestimation,":[26],"including":[27,124,144],"several":[28],"recent":[29],"ensemble":[30,58,65,88],"methods,":[31],"however":[32],"none":[33],"shown":[35],"success":[36],"sample-efficient":[38],"through":[40],"addressing":[41],"as":[44,64,104],"root":[46],"overestimation.":[49],"this":[51],"paper,":[52],"we":[53,84],"propose":[54],"MeanQ,":[55,123],"a":[56,105,132],"simple":[57],"method":[59],"that":[60,86],"estimates":[61],"values":[63],"means.":[66],"Despite":[67],"its":[68],"simplicity,":[69],"MeanQ":[70,139,162],"shows":[71],"remarkable":[72],"sample":[73,112],"efficiency":[74],"experiments":[76],"on":[77,160,176],"Atari":[79,137],"Learning":[80],"Environment":[81],"benchmark.":[82],"Importantly,":[83],"find":[85],"an":[87],"size":[90],"5":[91],"sufficiently":[92],"reduces":[93],"obviate":[97],"lagging":[99],"network,":[101],"eliminating":[102],"it":[103],"source":[106],"bias":[108],"further":[110],"gaining":[111],"efficiency.":[113],"We":[114],"justify":[115],"intuitively":[116],"empirically":[118],"design":[120],"choices":[121],"necessity":[126],"independent":[128],"experience":[129],"sampling.":[130],"On":[131],"set":[133],"26":[135],"benchmark":[136],"environments,":[138,156,172],"outperforms":[140,164],"all":[141],"tested":[142],"baselines,":[143],"best":[146],"available":[147,191],"baseline,":[148],"SUNRISE,":[149],"at":[150,167,192],"100K":[151],"interaction":[152,186],"steps":[153,169],"16/26":[155],"by":[158,174],"68%":[159],"average.":[161],"also":[163],"Rainbow":[165],"DQN":[166],"500K":[168],"21/26":[171],"49%":[175],"average,":[177],"achieves":[179],"average":[180],"human-level":[181],"performance":[182],"using":[183],"200K":[184],"($\\pm$100K)":[185],"steps.":[187],"Our":[188],"implementation":[189],"is":[190],"https://github.com/indylab/MeanQ.":[193]},"counts_by_year":[{"year":2024,"cited_by_count":4}],"updated_date":"2026-02-09T09:26:11.010843","created_date":"2025-10-10T00:00:00"}