{"id":"https://openalex.org/W3034214864","doi":"https://doi.org/10.24963/ijcai.2021/316","title":"Variational Model-based Policy Optimization","display_name":"Variational Model-based Policy Optimization","publication_year":2021,"publication_date":"2021-08-01","ids":{"openalex":"https://openalex.org/W3034214864","doi":"https://doi.org/10.24963/ijcai.2021/316","mag":"3034214864"},"language":"en","primary_location":{"id":"doi:10.24963/ijcai.2021/316","is_oa":true,"landing_page_url":"https://doi.org/10.24963/ijcai.2021/316","pdf_url":"https://www.ijcai.org/proceedings/2021/0316.pdf","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.ijcai.org/proceedings/2021/0316.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5039948454","display_name":"Yinlam Chow","orcid":"https://orcid.org/0000-0002-7533-8300"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Yinlam Chow","raw_affiliation_strings":["Google AI","Google,,,,,"],"affiliations":[{"raw_affiliation_string":"Google AI","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google,,,,,","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067671433","display_name":"Brandon Cui","orcid":null},"institutions":[{"id":"https://openalex.org/I2252078561","display_name":"Meta (Israel)","ror":"https://ror.org/02388em19","country_code":"IL","type":"company","lineage":["https://openalex.org/I2252078561","https://openalex.org/I4210114444"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Brandon Cui","raw_affiliation_strings":["Facebook AI Research","Facebook"],"affiliations":[{"raw_affiliation_string":"Facebook AI Research","institution_ids":["https://openalex.org/I2252078561"]},{"raw_affiliation_string":"Facebook","institution_ids":["https://openalex.org/I2252078561"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071567938","display_name":"Moonkyung Ryu","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Moonkyung Ryu","raw_affiliation_strings":["Google","Google,,,,,"],"affiliations":[{"raw_affiliation_string":"Google","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google,,,,,","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5013843778","display_name":"Mohammad Ghavamzadeh","orcid":"https://orcid.org/0000-0003-0930-8688"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mohammad Ghavamzadeh","raw_affiliation_strings":["Google Research","Google,,,,,"],"affiliations":[{"raw_affiliation_string":"Google Research","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google,,,,,","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5039948454"],"corresponding_institution_ids":["https://openalex.org/I1291425158"],"apc_list":null,"apc_paid":null,"fwci":0.4192,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.67002673,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"2292","last_page":"2299"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9998000264167786,"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":0.9998000264167786,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.996399998664856,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.991599977016449,"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/leverage","display_name":"Leverage (statistics)","score":0.6404641270637512},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6364742517471313},{"id":"https://openalex.org/keywords/maximization","display_name":"Maximization","score":0.6363449096679688},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.6202861070632935},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.564262866973877},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4340862035751343},{"id":"https://openalex.org/keywords/expectation\u2013maximization-algorithm","display_name":"Expectation\u2013maximization algorithm","score":0.4322042763233185},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.35826778411865234},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3339557647705078},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.22507312893867493},{"id":"https://openalex.org/keywords/maximum-likelihood","display_name":"Maximum likelihood","score":0.09723085165023804}],"concepts":[{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.6404641270637512},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6364742517471313},{"id":"https://openalex.org/C2776330181","wikidata":"https://www.wikidata.org/wiki/Q18358244","display_name":"Maximization","level":2,"score":0.6363449096679688},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6202861070632935},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.564262866973877},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4340862035751343},{"id":"https://openalex.org/C182081679","wikidata":"https://www.wikidata.org/wiki/Q1275153","display_name":"Expectation\u2013maximization algorithm","level":3,"score":0.4322042763233185},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35826778411865234},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3339557647705078},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.22507312893867493},{"id":"https://openalex.org/C49781872","wikidata":"https://www.wikidata.org/wiki/Q1045555","display_name":"Maximum likelihood","level":2,"score":0.09723085165023804},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.0}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.24963/ijcai.2021/316","is_oa":true,"landing_page_url":"https://doi.org/10.24963/ijcai.2021/316","pdf_url":"https://www.ijcai.org/proceedings/2021/0316.pdf","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2006.05443","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2006.05443","pdf_url":"https://arxiv.org/pdf/2006.05443","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":null,"raw_type":"text"},{"id":"mag:3034214864","is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2006.05443.pdf","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},{"id":"doi:10.48550/arxiv.2006.05443","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2006.05443","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"}],"best_oa_location":{"id":"doi:10.24963/ijcai.2021/316","is_oa":true,"landing_page_url":"https://doi.org/10.24963/ijcai.2021/316","pdf_url":"https://www.ijcai.org/proceedings/2021/0316.pdf","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3034214864.pdf","grobid_xml":"https://content.openalex.org/works/W3034214864.grobid-xml"},"referenced_works_count":33,"referenced_works":["https://openalex.org/W1499669280","https://openalex.org/W1596600792","https://openalex.org/W1757796397","https://openalex.org/W1771410628","https://openalex.org/W2080039641","https://openalex.org/W2098284983","https://openalex.org/W2098432798","https://openalex.org/W2107464055","https://openalex.org/W2107662876","https://openalex.org/W2109169869","https://openalex.org/W2121103318","https://openalex.org/W2121863487","https://openalex.org/W2139914196","https://openalex.org/W2145339207","https://openalex.org/W2147032798","https://openalex.org/W2155772159","https://openalex.org/W2166851633","https://openalex.org/W2169269897","https://openalex.org/W2211399972","https://openalex.org/W2296360731","https://openalex.org/W2529601334","https://openalex.org/W2593044849","https://openalex.org/W2799151646","https://openalex.org/W2890960027","https://openalex.org/W2914169655","https://openalex.org/W2962777832","https://openalex.org/W2962902376","https://openalex.org/W2963430173","https://openalex.org/W2963884015","https://openalex.org/W2970277495","https://openalex.org/W2970599228","https://openalex.org/W2995322606","https://openalex.org/W3007008575"],"related_works":["https://openalex.org/W3189406951","https://openalex.org/W2972448932","https://openalex.org/W2985997154","https://openalex.org/W2272837357","https://openalex.org/W2900677334","https://openalex.org/W2133120480","https://openalex.org/W2951201424","https://openalex.org/W2950182411","https://openalex.org/W2902113407","https://openalex.org/W2783076225","https://openalex.org/W2963177395","https://openalex.org/W3131864540","https://openalex.org/W2918175465","https://openalex.org/W128438430","https://openalex.org/W2950574573","https://openalex.org/W2963923417","https://openalex.org/W3080700717","https://openalex.org/W2990948232","https://openalex.org/W2570076534","https://openalex.org/W3188583492"],"abstract_inverted_index":{"Model-based":[0],"reinforcement":[1],"learning":[2],"(RL)":[3],"algorithms":[4,33,143,218],"allow":[5],"us":[6,97],"to":[7,22,53,57,98,159,194],"combine":[8],"model-generated":[9],"data":[10,25,42,48],"with":[11,16,211],"those":[12],"collected":[13],"from":[14],"interaction":[15],"the":[17,24,38,45,75,125,129,145,150,161,203],"real":[18],"system":[19],"in":[20,28,40,163],"order":[21],"alleviate":[23],"efficiency":[26,223],"problem":[27],"RL.":[29],"However,":[30],"designing":[31],"such":[32,84],"is":[34,56,189],"often":[35],"challenging":[36],"because":[37],"bias":[39],"simulated":[41],"may":[43],"overshadow":[44],"ease":[46],"of":[47,92,115,173],"generation.":[49],"A":[50],"potential":[51],"solution":[52],"this":[54,71],"challenge":[55],"jointly":[58],"learn":[59,110],"and":[60,63,79,82,103,109,118,137,147,192,215,219,224],"improve":[61],"model":[62,117],"policy":[64,108,120,127,139,186],"using":[65],"a":[66,89,93,106,111,116,119,164,171],"universal":[67],"objective":[68,86],"function.":[69],"In":[70],"paper,":[72],"we":[73,207],"leverage":[74],"connection":[76],"between":[77],"RL":[78,217],"probabilistic":[80],"inference,":[81],"formulate":[83],"an":[85],"function":[87],"as":[88],"variational":[90,112,131,151,184],"lower-bound":[91],"log-likelihood.":[94],"This":[95],"allows":[96],"use":[99],"expectation":[100],"maximization":[101],"(EM)":[102],"iteratively":[104],"fix":[105],"baseline":[107,126],"distribution,":[113],"consisting":[114],"(E-step),":[121],"followed":[122],"by":[123,154],"improving":[124],"given":[128],"learned":[130,153],"distribution":[132,152],"(M-step).":[133],"We":[134],"propose":[135],"model-based":[136,166,180,185,214],"model-free":[138,199,216],"iteration":[140],"(actor-critic)":[141],"style":[142],"for":[144],"E-step":[146],"show":[148,177,220],"how":[149],"them":[155],"can":[156],"be":[157],"used":[158],"optimize":[160],"M-step":[162],"fully":[165],"fashion.":[167],"Our":[168],"experiments":[169],"on":[170],"number":[172],"continuous":[174],"control":[175,205],"tasks":[176],"that":[178],"our":[179],"(E-step)":[181,200],"algorithm,":[182],"called":[183],"optimization":[187],"(VMBPO),":[188],"more":[190],"sample-efficient":[191],"robust":[193],"hyper-parameter":[195],"tuning":[196],"than":[197],"its":[198,221],"counterpart.":[201],"Using":[202],"same":[204],"tasks,":[206],"also":[208],"compare":[209],"VMBPO":[210],"several":[212],"state-of-the-art":[213],"sample":[222],"performance.":[225]},"counts_by_year":[{"year":2022,"cited_by_count":1},{"year":2021,"cited_by_count":2},{"year":2020,"cited_by_count":1}],"updated_date":"2026-03-04T09:10:02.777135","created_date":"2025-10-10T00:00:00"}