{"id":"https://openalex.org/W4402699008","doi":"https://doi.org/10.48550/arxiv.2408.12830","title":"Mitigating Distribution Shift in Model-based Offline RL via Shifts-aware Reward Learning","display_name":"Mitigating Distribution Shift in Model-based Offline RL via Shifts-aware Reward Learning","publication_year":2024,"publication_date":"2024-08-23","ids":{"openalex":"https://openalex.org/W4402699008","doi":"https://doi.org/10.48550/arxiv.2408.12830"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2408.12830","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2408.12830","pdf_url":"https://arxiv.org/pdf/2408.12830","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":null},"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2408.12830","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5003583859","display_name":"Wang Luo","orcid":"https://orcid.org/0000-0001-8230-5876"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Luo, Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100327709","display_name":"Haoran Li","orcid":"https://orcid.org/0000-0001-6421-0128"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Haoran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101797193","display_name":"Zicheng Zhang","orcid":"https://orcid.org/0000-0002-9553-7358"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Zicheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046469888","display_name":"Congying Han","orcid":"https://orcid.org/0000-0002-3445-4620"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Congying","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhou, Chi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Chi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004132894","display_name":"Jiayu Lv","orcid":"https://orcid.org/0000-0001-8703-4964"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lv, Jiayu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100561359","display_name":"Tiande Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Tiande","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5003583859"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.8059999942779541,"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":0.8059999942779541,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.8222016096115112},{"id":"https://openalex.org/keywords/reinforcement","display_name":"Reinforcement","score":0.5892926454544067},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5428375005722046},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.42636287212371826},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.29584959149360657},{"id":"https://openalex.org/keywords/social-psychology","display_name":"Social psychology","score":0.18508529663085938}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.8222016096115112},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.5892926454544067},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5428375005722046},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.42636287212371826},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.29584959149360657},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.18508529663085938}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2408.12830","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2408.12830","pdf_url":"https://arxiv.org/pdf/2408.12830","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":null},{"id":"doi:10.48550/arxiv.2408.12830","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2408.12830","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":"pmh:oai:arXiv.org:2408.12830","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2408.12830","pdf_url":"https://arxiv.org/pdf/2408.12830","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":null},"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/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W4310083477","https://openalex.org/W2328553770","https://openalex.org/W2920061524","https://openalex.org/W1977959518","https://openalex.org/W2038908348","https://openalex.org/W2107890255","https://openalex.org/W2106552856","https://openalex.org/W2145821588"],"abstract_inverted_index":{"Model-based":[0],"offline":[1],"reinforcement":[2],"learning":[3,115],"trains":[4],"policies":[5],"using":[6],"pre-collected":[7],"datasets":[8],"and":[9,39,50,72,77,87,116,154],"learned":[10],"environment":[11,19],"models,":[12],"eliminating":[13],"the":[14,64,109,134,148],"need":[15],"for":[16,137],"direct":[17],"real-world":[18],"interaction.":[20],"However,":[21],"this":[22,33],"paradigm":[23],"is":[24],"inherently":[25],"challenged":[26],"by":[27,35],"distribution":[28,152],"shift~(DS).":[29],"Existing":[30],"methods":[31],"address":[32],"issue":[34],"leveraging":[36],"off-policy":[37],"mechanisms":[38],"estimating":[40],"model":[41,70],"uncertainty,":[42],"but":[43],"they":[44],"often":[45],"result":[46],"in":[47],"inconsistent":[48],"objectives":[49],"lack":[51],"a":[52,59,97,102,125],"unified":[53,103],"theoretical":[54,76,162],"foundation.":[55],"This":[56],"paper":[57],"offers":[58],"comprehensive":[60],"analysis":[61],"that":[62,128,147],"disentangles":[63],"problem":[65],"into":[66],"two":[67],"fundamental":[68],"components:":[69],"bias":[71],"policy":[73,89,118,139],"shift.":[74],"Our":[75],"empirical":[78],"investigations":[79],"reveal":[80],"how":[81],"these":[82,93],"factors":[83],"distort":[84],"value":[85,114],"estimation":[86],"restrict":[88],"optimization.":[90,140],"To":[91],"tackle":[92],"challenges,":[94],"we":[95,123],"derive":[96],"novel":[98],"shifts-aware":[99],"reward":[100,111,136],"through":[101],"probabilistic":[104],"inference":[105],"framework,":[106],"which":[107],"modifies":[108],"vanilla":[110],"to":[112,132],"refine":[113],"facilitate":[117],"training.":[119],"Building":[120],"on":[121],"this,":[122],"develop":[124],"practical":[126],"implementation":[127],"leverages":[129],"classifier-based":[130],"techniques":[131],"approximate":[133],"adjusted":[135],"effective":[138],"Empirical":[141],"results":[142],"across":[143],"multiple":[144],"benchmarks":[145],"demonstrate":[146],"proposed":[149],"approach":[150],"mitigates":[151],"shift":[153],"achieves":[155],"superior":[156],"or":[157],"comparable":[158],"performance,":[159],"validating":[160],"our":[161],"insights.":[163]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-05-05T08:41:31.759640","created_date":"2024-09-21T00:00:00"}