{"id":"https://openalex.org/W4408355010","doi":"https://doi.org/10.1109/icassp49660.2025.10890751","title":"Improved Techniques for Offline Reinforcement Learning: Advantage Value Estimation and Layernorm","display_name":"Improved Techniques for Offline Reinforcement Learning: Advantage Value Estimation and Layernorm","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408355010","doi":"https://doi.org/10.1109/icassp49660.2025.10890751"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10890751","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10890751","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100754427","display_name":"Xiaosong Liu","orcid":"https://orcid.org/0000-0001-6158-5928"},"institutions":[{"id":"https://openalex.org/I3923682","display_name":"Soochow University","ror":"https://ror.org/05t8y2r12","country_code":"CN","type":"education","lineage":["https://openalex.org/I3923682"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Xiaosong Liu","raw_affiliation_strings":["Soochow University,Computer Science and Technology,Soochow,China"],"affiliations":[{"raw_affiliation_string":"Soochow University,Computer Science and Technology,Soochow,China","institution_ids":["https://openalex.org/I3923682"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100666897","display_name":"Quan Liu","orcid":"https://orcid.org/0000-0002-0941-8767"},"institutions":[{"id":"https://openalex.org/I3923682","display_name":"Soochow University","ror":"https://ror.org/05t8y2r12","country_code":"CN","type":"education","lineage":["https://openalex.org/I3923682"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Quan Liu","raw_affiliation_strings":["Soochow University,Computer Science and Technology,Soochow,China"],"affiliations":[{"raw_affiliation_string":"Soochow University,Computer Science and Technology,Soochow,China","institution_ids":["https://openalex.org/I3923682"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5108030925","display_name":"Lan Wu","orcid":"https://orcid.org/0000-0003-0892-6601"},"institutions":[{"id":"https://openalex.org/I3923682","display_name":"Soochow University","ror":"https://ror.org/05t8y2r12","country_code":"CN","type":"education","lineage":["https://openalex.org/I3923682"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lan Wu","raw_affiliation_strings":["Soochow University,Computer Science and Technology,Soochow,China"],"affiliations":[{"raw_affiliation_string":"Soochow University,Computer Science and Technology,Soochow,China","institution_ids":["https://openalex.org/I3923682"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5100754427"],"corresponding_institution_ids":["https://openalex.org/I3923682"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.07360952,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10328","display_name":"Supply Chain and Inventory Management","score":0.48750001192092896,"subfield":{"id":"https://openalex.org/subfields/1404","display_name":"Management Information Systems"},"field":{"id":"https://openalex.org/fields/14","display_name":"Business, Management and Accounting"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T10328","display_name":"Supply Chain and Inventory Management","score":0.48750001192092896,"subfield":{"id":"https://openalex.org/subfields/1404","display_name":"Management Information Systems"},"field":{"id":"https://openalex.org/fields/14","display_name":"Business, Management and Accounting"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.8246066570281982},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7036974430084229},{"id":"https://openalex.org/keywords/value","display_name":"Value (mathematics)","score":0.45556640625},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4474687874317169},{"id":"https://openalex.org/keywords/reinforcement","display_name":"Reinforcement","score":0.44295981526374817},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.36946040391921997},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.12963703274726868},{"id":"https://openalex.org/keywords/structural-engineering","display_name":"Structural engineering","score":0.08215618133544922}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.8246066570281982},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7036974430084229},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.45556640625},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4474687874317169},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.44295981526374817},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36946040391921997},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.12963703274726868},{"id":"https://openalex.org/C66938386","wikidata":"https://www.wikidata.org/wiki/Q633538","display_name":"Structural engineering","level":1,"score":0.08215618133544922}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10890751","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10890751","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":19,"referenced_works":["https://openalex.org/W2145339207","https://openalex.org/W4283590445","https://openalex.org/W4286253093","https://openalex.org/W4360584316","https://openalex.org/W4362672763","https://openalex.org/W4385245566","https://openalex.org/W6748839928","https://openalex.org/W6757469721","https://openalex.org/W6776438516","https://openalex.org/W6776601253","https://openalex.org/W6779265984","https://openalex.org/W6796589144","https://openalex.org/W6802659552","https://openalex.org/W6810030924","https://openalex.org/W6838356327","https://openalex.org/W6840893754","https://openalex.org/W6846345420","https://openalex.org/W6849433528","https://openalex.org/W6850886341"],"related_works":["https://openalex.org/W2961085424","https://openalex.org/W4306674287","https://openalex.org/W4387369504","https://openalex.org/W3046775127","https://openalex.org/W4394896187","https://openalex.org/W3170094116","https://openalex.org/W4386462264","https://openalex.org/W3107602296","https://openalex.org/W4364306694","https://openalex.org/W4312192474"],"abstract_inverted_index":{"Offline":[0],"reinforcement":[1],"learning,":[2],"which":[3,30,77],"aims":[4],"to":[5,17,119],"learn":[6],"an":[7],"optimal":[8],"policy":[9],"from":[10],"a":[11,45,49],"previously":[12],"collected":[13],"static":[14],"datasets.":[15,142],"Due":[16],"the":[18,32,36,66,80,85,90,96,103,112],"overestimation":[19,97],"caused":[20],"by":[21],"extrapolation":[22],"error,":[23],"offline":[24],"algorithms":[25],"adopt":[26],"overly":[27],"pessimistic":[28],"approaches,":[29],"compromise":[31],"generalization":[33,135],"ability":[34,136],"of":[35,68,98],"learned":[37],"policy.":[38],"To":[39],"address":[40],"these":[41],"issues,":[42],"we":[43],"propose":[44],"method":[46],"that":[47],"adopts":[48],"mild":[50],"constraint":[51],"learning":[52,83],"approach,":[53],"comprising":[54],"two":[55],"components:":[56],"advantage":[57],"value":[58,67,91],"estimation":[59],"and":[60,70,84,101,140],"layernorm.":[61],"The":[62],"first":[63],"component":[64],"estimates":[65],"actions":[69,74,100],"then":[71],"selects":[72],"valuable":[73],"for":[75],"imitation,":[76],"moderately":[78],"relaxes":[79],"strict":[81],"conservative":[82],"second":[86],"applies":[87],"layernorm":[88],"in":[89,111,127],"function":[92],"network,":[93],"effectively":[94],"addressing":[95],"out-of-distribution(OOD)":[99],"stabilizing":[102],"training":[104],"process.":[105],"Experimental":[106],"results":[107],"on":[108,137],"various":[109],"tasks":[110],"D4RL":[113],"MuJoCo":[114],"benchmark":[115],"show":[116],"that,":[117],"compared":[118],"baseline":[120],"methods,":[121],"our":[122,131],"algorithm":[123,132],"achieves":[124],"better":[125],"performance":[126],"most":[128],"tasks.":[129],"Especially,":[130],"exhibits":[133],"well":[134],"random,":[138],"medium-replay,":[139],"full-replay":[141]},"counts_by_year":[],"updated_date":"2025-12-21T01:58:51.020947","created_date":"2025-10-10T00:00:00"}