{"id":"https://openalex.org/W2944187456","doi":"https://doi.org/10.1007/s10994-022-06232-6","title":"Smoothing policies and safe policy gradients","display_name":"Smoothing policies and safe policy gradients","publication_year":2022,"publication_date":"2022-10-20","ids":{"openalex":"https://openalex.org/W2944187456","doi":"https://doi.org/10.1007/s10994-022-06232-6","mag":"2944187456"},"language":"en","primary_location":{"id":"doi:10.1007/s10994-022-06232-6","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s10994-022-06232-6","pdf_url":"https://link.springer.com/content/pdf/10.1007/s10994-022-06232-6.pdf","source":{"id":"https://openalex.org/S62148650","display_name":"Machine Learning","issn_l":"0885-6125","issn":["0885-6125","1573-0565"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Machine Learning","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://link.springer.com/content/pdf/10.1007/s10994-022-06232-6.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5015303026","display_name":"Matteo Papini","orcid":"https://orcid.org/0000-0002-3807-3171"},"institutions":[{"id":"https://openalex.org/I170486558","display_name":"Universitat Pompeu Fabra","ror":"https://ror.org/04n0g0b29","country_code":"ES","type":"education","lineage":["https://openalex.org/I170486558"]}],"countries":["ES"],"is_corresponding":true,"raw_author_name":"Matteo Papini","raw_affiliation_strings":["Universitat Pompeu Fabra, Barcelona, Spain"],"raw_orcid":"https://orcid.org/0000-0002-3807-3171","affiliations":[{"raw_affiliation_string":"Universitat Pompeu Fabra, Barcelona, Spain","institution_ids":["https://openalex.org/I170486558"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5091526684","display_name":"Matteo Pirotta","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Matteo Pirotta","raw_affiliation_strings":["Meta, Paris, France"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Meta, Paris, France","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5017130830","display_name":"Marcello Restelli","orcid":"https://orcid.org/0000-0002-6322-1076"},"institutions":[{"id":"https://openalex.org/I93860229","display_name":"Politecnico di Milano","ror":"https://ror.org/01nffqt88","country_code":"IT","type":"education","lineage":["https://openalex.org/I93860229"]}],"countries":["IT"],"is_corresponding":false,"raw_author_name":"Marcello Restelli","raw_affiliation_strings":["Politecnico di Milano, Milan, Italy"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Politecnico di Milano, Milan, Italy","institution_ids":["https://openalex.org/I93860229"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5015303026"],"corresponding_institution_ids":["https://openalex.org/I170486558"],"apc_list":{"value":2890,"currency":"USD","value_usd":2890},"apc_paid":{"value":2890,"currency":"USD","value_usd":2890},"fwci":1.2839,"has_fulltext":true,"cited_by_count":22,"citation_normalized_percentile":{"value":0.82269425,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":98},"biblio":{"volume":"111","issue":"11","first_page":"4081","last_page":"4137"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9988999962806702,"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":0.9988999962806702,"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/T12814","display_name":"Gaussian Processes and Bayesian Inference","score":0.9873999953269958,"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/T11099","display_name":"Autonomous Vehicle Technology and Safety","score":0.9801999926567078,"subfield":{"id":"https://openalex.org/subfields/2203","display_name":"Automotive Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.739666759967804},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6622620820999146},{"id":"https://openalex.org/keywords/estimator","display_name":"Estimator","score":0.6448206901550293},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.582710862159729},{"id":"https://openalex.org/keywords/smoothing","display_name":"Smoothing","score":0.503546416759491},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.4899868071079254},{"id":"https://openalex.org/keywords/gaussian-process","display_name":"Gaussian process","score":0.4808129072189331},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4629479646682739},{"id":"https://openalex.org/keywords/monotonic-function","display_name":"Monotonic function","score":0.4309481978416443},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.3854641318321228},{"id":"https://openalex.org/keywords/gaussian","display_name":"Gaussian","score":0.3623286485671997},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.22227898240089417},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.1119515597820282}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.739666759967804},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6622620820999146},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.6448206901550293},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.582710862159729},{"id":"https://openalex.org/C3770464","wikidata":"https://www.wikidata.org/wiki/Q775963","display_name":"Smoothing","level":2,"score":0.503546416759491},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.4899868071079254},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.4808129072189331},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4629479646682739},{"id":"https://openalex.org/C72169020","wikidata":"https://www.wikidata.org/wiki/Q194404","display_name":"Monotonic function","level":2,"score":0.4309481978416443},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3854641318321228},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.3623286485671997},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.22227898240089417},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.1119515597820282},{"id":"https://openalex.org/C121955636","wikidata":"https://www.wikidata.org/wiki/Q4116214","display_name":"Accounting","level":1,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","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}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1007/s10994-022-06232-6","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s10994-022-06232-6","pdf_url":"https://link.springer.com/content/pdf/10.1007/s10994-022-06232-6.pdf","source":{"id":"https://openalex.org/S62148650","display_name":"Machine Learning","issn_l":"0885-6125","issn":["0885-6125","1573-0565"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Machine Learning","raw_type":"journal-article"},{"id":"pmh:oai:repositori-api.upf.edu:10230/56240","is_oa":true,"landing_page_url":"http://hdl.handle.net/10230/56240","pdf_url":null,"source":{"id":"https://openalex.org/S4306402615","display_name":"Repositori digital de la UPF (Universitat Pompeu Fabra)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I170486558","host_organization_name":"Universitat Pompeu Fabra","host_organization_lineage":["https://openalex.org/I170486558"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/article"},{"id":"pmh:oai:air.unimi.it:2434/1225915","is_oa":true,"landing_page_url":"https://hdl.handle.net/2434/1225915","pdf_url":null,"source":{"id":"https://openalex.org/S4306400516","display_name":"Archivio Istituzionale della Ricerca (Universita Degli Studi Di Milano)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I189158943","host_organization_name":"University of Milan","host_organization_lineage":["https://openalex.org/I189158943"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/article"},{"id":"pmh:oai:re.public.polimi.it:11311/1231794","is_oa":true,"landing_page_url":"https://hdl.handle.net/11311/1231794","pdf_url":null,"source":{"id":"https://openalex.org/S4306400312","display_name":"Virtual Community of Pathological Anatomy (University of Castilla La Mancha)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I79189158","host_organization_name":"University of Castilla-La Mancha","host_organization_lineage":["https://openalex.org/I79189158"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"info:eu-repo/semantics/article"}],"best_oa_location":{"id":"doi:10.1007/s10994-022-06232-6","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s10994-022-06232-6","pdf_url":"https://link.springer.com/content/pdf/10.1007/s10994-022-06232-6.pdf","source":{"id":"https://openalex.org/S62148650","display_name":"Machine Learning","issn_l":"0885-6125","issn":["0885-6125","1573-0565"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Machine Learning","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320318602","display_name":"Universitat Pompeu Fabra","ror":"https://ror.org/04n0g0b29"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2944187456.pdf","grobid_xml":"https://content.openalex.org/works/W2944187456.grobid-xml"},"referenced_works_count":117,"referenced_works":["https://openalex.org/W1497976081","https://openalex.org/W1498636830","https://openalex.org/W1522301498","https://openalex.org/W1529558080","https://openalex.org/W1575592356","https://openalex.org/W1606011487","https://openalex.org/W1634042229","https://openalex.org/W1786332878","https://openalex.org/W1845972764","https://openalex.org/W1850488217","https://openalex.org/W1977655452","https://openalex.org/W1994616650","https://openalex.org/W2012587148","https://openalex.org/W2014583745","https://openalex.org/W2046859786","https://openalex.org/W2054226461","https://openalex.org/W2091565802","https://openalex.org/W2094387729","https://openalex.org/W2096727211","https://openalex.org/W2097568041","https://openalex.org/W2101075098","https://openalex.org/W2112964839","https://openalex.org/W2114537044","https://openalex.org/W2115126871","https://openalex.org/W2116383462","https://openalex.org/W2119567691","https://openalex.org/W2119717200","https://openalex.org/W2121863487","https://openalex.org/W2124477018","https://openalex.org/W2125612430","https://openalex.org/W2134491302","https://openalex.org/W2142426721","https://openalex.org/W2142899754","https://openalex.org/W2144446635","https://openalex.org/W2145339207","https://openalex.org/W2155027007","https://openalex.org/W2156718681","https://openalex.org/W2165150801","https://openalex.org/W2169619645","https://openalex.org/W2288565641","https://openalex.org/W2342662072","https://openalex.org/W2427917354","https://openalex.org/W2462906003","https://openalex.org/W2552231430","https://openalex.org/W2604272474","https://openalex.org/W2618318883","https://openalex.org/W2619268125","https://openalex.org/W2734657980","https://openalex.org/W2740828027","https://openalex.org/W2749928749","https://openalex.org/W2753569380","https://openalex.org/W2767002724","https://openalex.org/W2781726626","https://openalex.org/W2783932892","https://openalex.org/W2786036274","https://openalex.org/W2786303200","https://openalex.org/W2788729564","https://openalex.org/W2788837839","https://openalex.org/W2798273187","https://openalex.org/W2803543472","https://openalex.org/W2804329956","https://openalex.org/W2822752092","https://openalex.org/W2885163910","https://openalex.org/W2902907165","https://openalex.org/W2913773024","https://openalex.org/W2918175465","https://openalex.org/W2948432982","https://openalex.org/W2949608212","https://openalex.org/W2950650380","https://openalex.org/W2952720101","https://openalex.org/W2963184939","https://openalex.org/W2963400359","https://openalex.org/W2964043796","https://openalex.org/W2964108826","https://openalex.org/W2964986650","https://openalex.org/W2974544365","https://openalex.org/W2991598122","https://openalex.org/W2998050631","https://openalex.org/W2998089193","https://openalex.org/W3005438958","https://openalex.org/W3009440920","https://openalex.org/W3035020089","https://openalex.org/W3037603015","https://openalex.org/W3093206925","https://openalex.org/W3093528669","https://openalex.org/W3103182070","https://openalex.org/W3104139512","https://openalex.org/W3106238320","https://openalex.org/W3171937521","https://openalex.org/W3186110329","https://openalex.org/W4205513846","https://openalex.org/W4214717370","https://openalex.org/W4241536761","https://openalex.org/W4246388106","https://openalex.org/W4388297583","https://openalex.org/W6600644339","https://openalex.org/W6629893325","https://openalex.org/W6629916206","https://openalex.org/W6631151876","https://openalex.org/W6638018090","https://openalex.org/W6640490175","https://openalex.org/W6644363735","https://openalex.org/W6676516282","https://openalex.org/W6676898700","https://openalex.org/W6677916085","https://openalex.org/W6679257226","https://openalex.org/W6682367392","https://openalex.org/W6683195989","https://openalex.org/W6729785568","https://openalex.org/W6737893269","https://openalex.org/W6743613440","https://openalex.org/W6747790125","https://openalex.org/W6751972096","https://openalex.org/W6759014971","https://openalex.org/W6762029327","https://openalex.org/W6780559895","https://openalex.org/W7001894244"],"related_works":["https://openalex.org/W4380682190","https://openalex.org/W2383807498","https://openalex.org/W1978572805","https://openalex.org/W1997992934","https://openalex.org/W1987225439","https://openalex.org/W4315701745","https://openalex.org/W1990290471","https://openalex.org/W4238188170","https://openalex.org/W2945307361","https://openalex.org/W2125114371"],"abstract_inverted_index":{"Abstract":[0],"Policy":[1],"gradient":[2,161],"(PG)":[3],"algorithms":[4],"are":[5,67,147],"among":[6],"the":[7,11,25,35,75,87,125,148,152,156,160],"best":[8],"candidates":[9],"for":[10,105],"much-anticipated":[12],"applications":[13],"of":[14,28,50,90,109,127,151,159,168],"reinforcement":[15],"learning":[16,36,76],"to":[17,80,132],"real-world":[18],"control":[19],"tasks,":[20],"such":[21],"as":[22,86],"robotics.":[23],"However,":[24],"trial-and-error":[26],"nature":[27],"these":[29,169],"methods":[30],"poses":[31],"safety":[32,60],"issues":[33],"whenever":[34],"process":[37],"itself":[38],"must":[39],"be":[40],"performed":[41],"on":[42,115,124],"a":[43,58,70,97,106,164,173],"physical":[44],"system":[45],"or":[46],"involves":[47],"any":[48],"form":[49],"human-computer":[51],"interaction.":[52],"In":[53],"this":[54],"paper,":[55],"we":[56,101,171],"address":[57],"specific":[59],"formulation,":[61],"where":[62],"both":[63],"goals":[64],"and":[65,74,155],"dangers":[66],"encoded":[68],"in":[69],"scalar":[71],"reward":[72],"signal":[73],"agent":[77],"is":[78],"constrained":[79],"never":[81],"worsen":[82],"its":[83],"performance,":[84],"measured":[85],"expected":[88],"sum":[89],"rewards.":[91],"By":[92],"studying":[93],"actor-only":[94],"PG":[95,128,174],"from":[96],"stochastic":[98],"optimization":[99],"perspective,":[100],"establish":[102],"improvement":[103,139,178],"guarantees":[104],"wide":[107],"class":[108],"parametric":[110],"policies,":[111],"generalizing":[112],"existing":[113],"results":[114],"Gaussian":[116],"policies.":[117],"This,":[118],"together":[119],"with":[120,140,176],"novel":[121],"upper":[122],"bounds":[123],"variance":[126],"estimators,":[129],"allows":[130],"us":[131],"identify":[133],"meta-parameter":[134],"schedules":[135],"that":[136],"guarantee":[137],"monotonic":[138,177],"high":[141],"probability.":[142],"The":[143],"two":[144],"key":[145],"meta-parameters":[146],"step":[149],"size":[150,158],"parameter":[153],"updates":[154],"batch":[157],"estimates.":[162],"Through":[163],"joint,":[165],"adaptive":[166],"selection":[167],"meta-parameters,":[170],"obtain":[172],"algorithm":[175],"guarantees.":[179]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":4},{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":1},{"year":2021,"cited_by_count":5},{"year":2020,"cited_by_count":4},{"year":2019,"cited_by_count":1}],"updated_date":"2026-08-28T12:50:07.497085","created_date":"2025-10-10T00:00:00"}