{"id":"https://openalex.org/W4410040027","doi":"https://doi.org/10.1145/3733592","title":"Settling the Sample Complexity of Online Reinforcement Learning","display_name":"Settling the Sample Complexity of Online Reinforcement Learning","publication_year":2025,"publication_date":"2025-05-02","ids":{"openalex":"https://openalex.org/W4410040027","doi":"https://doi.org/10.1145/3733592"},"language":"en","primary_location":{"id":"doi:10.1145/3733592","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3733592","pdf_url":null,"source":{"id":"https://openalex.org/S118992489","display_name":"Journal of the ACM","issn_l":"0004-5411","issn":["0004-5411","1557-735X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of the ACM","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1145/3733592","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5065287257","display_name":"Zihan Zhang","orcid":"https://orcid.org/0009-0009-3379-9545"},"institutions":[{"id":"https://openalex.org/I20089843","display_name":"Princeton University","ror":"https://ror.org/00hx57361","country_code":"US","type":"education","lineage":["https://openalex.org/I20089843"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Zihan Zhang","raw_affiliation_strings":["Department of Electrical and Computer Engineering, Princeton University, Princeton, United States"],"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering, Princeton University, Princeton, United States","institution_ids":["https://openalex.org/I20089843"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100416078","display_name":"Yuxin Chen","orcid":"https://orcid.org/0000-0001-9256-5815"},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yuxin Chen","raw_affiliation_strings":["Department of Statistics and Data Science, University of Pennsylvania, Philadelphia, United States"],"affiliations":[{"raw_affiliation_string":"Department of Statistics and Data Science, University of Pennsylvania, Philadelphia, United States","institution_ids":["https://openalex.org/I79576946"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059740024","display_name":"Jason D. Lee","orcid":"https://orcid.org/0000-0003-0064-7800"},"institutions":[{"id":"https://openalex.org/I20089843","display_name":"Princeton University","ror":"https://ror.org/00hx57361","country_code":"US","type":"education","lineage":["https://openalex.org/I20089843"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jason Lee","raw_affiliation_strings":["Department of Electrical and Computer Engineering, Princeton University, Princeton, United States"],"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering, Princeton University, Princeton, United States","institution_ids":["https://openalex.org/I20089843"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5033061754","display_name":"Simon S. Du","orcid":"https://orcid.org/0000-0003-0056-8299"},"institutions":[{"id":"https://openalex.org/I201448701","display_name":"University of Washington","ror":"https://ror.org/00cvxb145","country_code":"US","type":"education","lineage":["https://openalex.org/I201448701"]},{"id":"https://openalex.org/I58610484","display_name":"Seattle University","ror":"https://ror.org/02jqc0m91","country_code":"US","type":"education","lineage":["https://openalex.org/I58610484"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Simon S. Du","raw_affiliation_strings":["Paul G. Allen School of Computer Science and Engineering, University of Washington, Seattle, United States"],"affiliations":[{"raw_affiliation_string":"Paul G. Allen School of Computer Science and Engineering, University of Washington, Seattle, United States","institution_ids":["https://openalex.org/I201448701","https://openalex.org/I58610484"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5065287257"],"corresponding_institution_ids":["https://openalex.org/I20089843"],"apc_list":null,"apc_paid":null,"fwci":5.7199,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.95302276,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":96},"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9954000115394592,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9954000115394592,"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/T12101","display_name":"Advanced Bandit Algorithms Research","score":0.9848999977111816,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T12288","display_name":"Optimization and Search Problems","score":0.97079998254776,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/settling","display_name":"Settling","score":0.741712749004364},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.732113242149353},{"id":"https://openalex.org/keywords/sample-complexity","display_name":"Sample complexity","score":0.6321164965629578},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.6284850835800171},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6147441267967224},{"id":"https://openalex.org/keywords/reinforcement","display_name":"Reinforcement","score":0.5517892837524414},{"id":"https://openalex.org/keywords/settling-time","display_name":"Settling time","score":0.47032177448272705},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3622361123561859},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.18739277124404907},{"id":"https://openalex.org/keywords/social-psychology","display_name":"Social psychology","score":0.12040796875953674},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.10488605499267578},{"id":"https://openalex.org/keywords/control-engineering","display_name":"Control engineering","score":0.08349859714508057},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.06778904795646667}],"concepts":[{"id":"https://openalex.org/C173209714","wikidata":"https://www.wikidata.org/wiki/Q923906","display_name":"Settling","level":2,"score":0.741712749004364},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.732113242149353},{"id":"https://openalex.org/C2778445095","wikidata":"https://www.wikidata.org/wiki/Q18354077","display_name":"Sample complexity","level":2,"score":0.6321164965629578},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.6284850835800171},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6147441267967224},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.5517892837524414},{"id":"https://openalex.org/C14781684","wikidata":"https://www.wikidata.org/wiki/Q3983320","display_name":"Settling time","level":3,"score":0.47032177448272705},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3622361123561859},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.18739277124404907},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.12040796875953674},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.10488605499267578},{"id":"https://openalex.org/C133731056","wikidata":"https://www.wikidata.org/wiki/Q4917288","display_name":"Control engineering","level":1,"score":0.08349859714508057},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.06778904795646667},{"id":"https://openalex.org/C160030872","wikidata":"https://www.wikidata.org/wiki/Q2142864","display_name":"Step response","level":2,"score":0.0},{"id":"https://openalex.org/C87717796","wikidata":"https://www.wikidata.org/wiki/Q146326","display_name":"Environmental engineering","level":1,"score":0.0},{"id":"https://openalex.org/C97355855","wikidata":"https://www.wikidata.org/wiki/Q11473","display_name":"Thermodynamics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3733592","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3733592","pdf_url":null,"source":{"id":"https://openalex.org/S118992489","display_name":"Journal of the ACM","issn_l":"0004-5411","issn":["0004-5411","1557-735X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of the ACM","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1145/3733592","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3733592","pdf_url":null,"source":{"id":"https://openalex.org/S118992489","display_name":"Journal of the ACM","issn_l":"0004-5411","issn":["0004-5411","1557-735X"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Journal of the ACM","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":27,"referenced_works":["https://openalex.org/W1583837637","https://openalex.org/W1850488217","https://openalex.org/W1988526405","https://openalex.org/W2124352385","https://openalex.org/W2161966552","https://openalex.org/W2591957553","https://openalex.org/W2769648743","https://openalex.org/W2912485311","https://openalex.org/W3009585715","https://openalex.org/W3020125231","https://openalex.org/W3034039613","https://openalex.org/W3090442079","https://openalex.org/W3094472530","https://openalex.org/W3102674085","https://openalex.org/W3138939485","https://openalex.org/W3182614517","https://openalex.org/W3207876914","https://openalex.org/W4224902895","https://openalex.org/W4281897969","https://openalex.org/W4287102143","https://openalex.org/W4287259025","https://openalex.org/W4287326219","https://openalex.org/W4292945870","https://openalex.org/W4366692289","https://openalex.org/W4385338412","https://openalex.org/W4392591475","https://openalex.org/W6677826109"],"related_works":["https://openalex.org/W1431544135","https://openalex.org/W2079208387","https://openalex.org/W2315908612","https://openalex.org/W2315186994","https://openalex.org/W2087757064","https://openalex.org/W2080005519","https://openalex.org/W2124144567","https://openalex.org/W2116157560","https://openalex.org/W2877093712","https://openalex.org/W4310614650"],"abstract_inverted_index":{"A":[0,129],"central":[1],"issue":[2],"lying":[3],"at":[4],"the":[5,28,106,125,131,137,143,151,156,179,199,209,215,244,252,258],"heart":[6],"of":[7,18,30,87,108,127,133,146,159,181,187,211,254],"online":[8,26,255],"reinforcement":[9],"learning":[10],"(RL)":[11],"is":[12,33,124,130,136,142,196],"data":[13],"efficiency.":[14],"While":[15],"a":[16,37,84,103,174,227,233,248],"number":[17,126,132,145,180],"recent":[19],"works":[20],"achieved":[21],"asymptotically":[22],"minimal":[23],"regret":[24,56,104,149],"in":[25,36,44,67,226,257],"RL,":[27],"optimality":[29],"these":[31],"results":[32],"only":[34],"guaranteed":[35],"\u201clarge-sample\u201d":[38],"regime,":[39],"imposing":[40],"enormous":[41],"burn-in":[42,60,168],"cost":[43,61],"order":[45,107],"for":[46,74,155,198],"their":[47],"algorithms":[48],"to":[49,53,173,184,192,207,238],"operate":[50],"optimally.":[51],"How":[52],"achieve":[54],"minimax-optimal":[55,197],"without":[57],"incurring":[58],"any":[59,167],"has":[62],"been":[63],"an":[64,92],"open":[65],"problem":[66,73],"RL":[68,256],"theory.":[69],"We":[70],"settle":[71],"this":[72],"finite-horizon":[75],"inhomogeneous":[76],"Markov":[77],"decision":[78],"processes.":[79],"Specifically,":[80],"we":[81,203],"prove":[82],"that":[83],"modified":[85],"version":[86],"MVP":[88],"(Monotonic":[89],"Value":[90],"Propagation),":[91],"optimistic":[93],"model-based":[94],"algorithm":[95],"proposed":[96],"by":[97],"Zhang":[98],"et":[99],"al.":[100],"[82],":[101],"achieves":[102],"on":[105,232],"(modulo":[109],"log":[110,193],"factors)":[111],"\\begin{equation*}":[112],"\\min":[113],"\\big":[114,119],"\\lbrace":[115],"\\sqrt":[116],"{SAH^3K},":[117],"\\,HK":[118],"\\rbrace,":[120],"\\end{equation*}":[121],"where":[122],"S":[123],"states,":[128],"actions,":[134],"H":[135],"horizon":[138],"length,":[139],"and":[140,218],"K":[141,162],"total":[144],"episodes.":[147],"This":[148],"matches":[150],"minimax":[152],"lower":[153],"bound":[154],"entire":[157],"range":[158],"sample":[160,176,245],"size":[161],"\u2265":[163],"1,":[164],"essentially":[165],"eliminating":[166],"requirement.":[169],"It":[170],"also":[171],"translates":[172],"PAC":[175],"complexity":[177],"(i.e.,":[178],"episodes":[182],"needed":[183],"yield":[185],"\u03b5-accuracy)":[186],"\\(\\frac{SAH^3}{\\varepsilon":[188],"^2}":[189],"\\)":[190],"up":[191],"factor,":[194],"which":[195],"full":[200],"\u03b5-range.":[201],"Further,":[202],"extend":[204],"our":[205],"theory":[206],"unveil":[208],"influences":[210],"problem-dependent":[212],"quantities":[213],"like":[214],"optimal":[216],"value/cost":[217],"certain":[219],"variances.":[220],"The":[221],"key":[222],"technical":[223],"innovation":[224],"lies":[225],"novel":[228],"analysis":[229,253],"paradigm":[230],"(based":[231],"new":[234],"concept":[235],"called":[236],"\u201cprofiles\u201d)":[237],"decouple":[239],"complicated":[240],"statistical":[241],"dependency":[242],"across":[243],"trajectories":[246],"\u2014":[247],"long-standing":[249],"challenge":[250],"facing":[251],"sample-starved":[259],"regime.":[260]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}