{"id":"https://openalex.org/W4408897262","doi":"https://doi.org/10.1109/icmlc63072.2024.10935257","title":"The Bifurcation Method: White-Box Observation Perturbation Attacks on Reinforcement Learning Agents on a Cyber Physical System","display_name":"The Bifurcation Method: White-Box Observation Perturbation Attacks on Reinforcement Learning Agents on a Cyber Physical System","publication_year":2024,"publication_date":"2024-09-20","ids":{"openalex":"https://openalex.org/W4408897262","doi":"https://doi.org/10.1109/icmlc63072.2024.10935257"},"language":"en","primary_location":{"id":"doi:10.1109/icmlc63072.2024.10935257","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icmlc63072.2024.10935257","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 International Conference on Machine Learning and Cybernetics (ICMLC)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5100493798","display_name":"Kiernan Broda-Milian","orcid":null},"institutions":[{"id":"https://openalex.org/I51768193","display_name":"Royal Military College of Canada","ror":"https://ror.org/04yr71909","country_code":"CA","type":"education","lineage":["https://openalex.org/I51768193"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"KIERNAN BRODA-MILIAN","raw_affiliation_strings":["Royal Military College of Canada, Electrical and Computer Engineering,Kingston,Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Royal Military College of Canada, Electrical and Computer Engineering,Kingston,Canada","institution_ids":["https://openalex.org/I51768193"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000287974","display_name":"Hanane Dagdougui","orcid":"https://orcid.org/0000-0001-7500-0468"},"institutions":[{"id":"https://openalex.org/I45683168","display_name":"Polytechnique Montr\u00e9al","ror":"https://ror.org/05f8d4e86","country_code":"CA","type":"education","lineage":["https://openalex.org/I45683168"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"HANANE DAGDOUGUI","raw_affiliation_strings":["Polytechnique Montreal, Computer and Software Engineering,Quebec,Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Polytechnique Montreal, Computer and Software Engineering,Quebec,Canada","institution_ids":["https://openalex.org/I45683168"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5020513405","display_name":"Ranwa Al Mallah","orcid":"https://orcid.org/0000-0003-3703-9729"},"institutions":[{"id":"https://openalex.org/I51768193","display_name":"Royal Military College of Canada","ror":"https://ror.org/04yr71909","country_code":"CA","type":"education","lineage":["https://openalex.org/I51768193"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"RANWA AL MALLAH","raw_affiliation_strings":["Royal Military College of Canada, Electrical and Computer Engineering,Kingston,Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Royal Military College of Canada, Electrical and Computer Engineering,Kingston,Canada","institution_ids":["https://openalex.org/I51768193"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.8877,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.78669695,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"571","last_page":"578"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T14420","display_name":"Advanced Research in Systems and Signal Processing","score":0.6227999925613403,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T14420","display_name":"Advanced Research in Systems and Signal Processing","score":0.6227999925613403,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10917","display_name":"Smart Grid Security and Resilience","score":0.619700014591217,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems 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/white-box","display_name":"White box","score":0.7714778184890747},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.7560325860977173},{"id":"https://openalex.org/keywords/bifurcation","display_name":"Bifurcation","score":0.7474714517593384},{"id":"https://openalex.org/keywords/cyber-physical-system","display_name":"Cyber-physical system","score":0.5758235454559326},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.515154242515564},{"id":"https://openalex.org/keywords/perturbation","display_name":"Perturbation (astronomy)","score":0.5132708549499512},{"id":"https://openalex.org/keywords/reinforcement","display_name":"Reinforcement","score":0.5114516019821167},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.40286052227020264},{"id":"https://openalex.org/keywords/control-theory","display_name":"Control theory (sociology)","score":0.38250765204429626},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.2842901051044464},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.24275165796279907},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.2042425572872162},{"id":"https://openalex.org/keywords/structural-engineering","display_name":"Structural engineering","score":0.15594875812530518},{"id":"https://openalex.org/keywords/nonlinear-system","display_name":"Nonlinear system","score":0.1486763060092926},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.11465582251548767},{"id":"https://openalex.org/keywords/quantum-mechanics","display_name":"Quantum mechanics","score":0.09359276294708252}],"concepts":[{"id":"https://openalex.org/C180932941","wikidata":"https://www.wikidata.org/wiki/Q997233","display_name":"White box","level":2,"score":0.7714778184890747},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7560325860977173},{"id":"https://openalex.org/C2781349735","wikidata":"https://www.wikidata.org/wiki/Q4904824","display_name":"Bifurcation","level":3,"score":0.7474714517593384},{"id":"https://openalex.org/C179768478","wikidata":"https://www.wikidata.org/wiki/Q1120057","display_name":"Cyber-physical system","level":2,"score":0.5758235454559326},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.515154242515564},{"id":"https://openalex.org/C177918212","wikidata":"https://www.wikidata.org/wiki/Q803623","display_name":"Perturbation (astronomy)","level":2,"score":0.5132708549499512},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.5114516019821167},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40286052227020264},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.38250765204429626},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.2842901051044464},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.24275165796279907},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2042425572872162},{"id":"https://openalex.org/C66938386","wikidata":"https://www.wikidata.org/wiki/Q633538","display_name":"Structural engineering","level":1,"score":0.15594875812530518},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.1486763060092926},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.11465582251548767},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.09359276294708252},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/icmlc63072.2024.10935257","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icmlc63072.2024.10935257","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 International Conference on Machine Learning and Cybernetics (ICMLC)","raw_type":"proceedings-article"},{"id":"pmh:oai:publications.polymtl.ca:64702","is_oa":false,"landing_page_url":"https://publications.polymtl.ca/64702/","pdf_url":null,"source":{"id":"https://openalex.org/S4306401013","display_name":"PolyPublie (\u00c9cole Polytechnique de Montr\u00e9al)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I45683168","host_organization_name":"Polytechnique Montr\u00e9al","host_organization_lineage":["https://openalex.org/I45683168"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":null,"raw_type":"Communication de conf\u00e9rence"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":11,"referenced_works":["https://openalex.org/W2545258053","https://openalex.org/W2951360122","https://openalex.org/W3108288031","https://openalex.org/W3112388352","https://openalex.org/W3200885897","https://openalex.org/W4205346763","https://openalex.org/W4289792178","https://openalex.org/W4386362892","https://openalex.org/W6752654261","https://openalex.org/W6764982603","https://openalex.org/W6784827930"],"related_works":["https://openalex.org/W3004173571","https://openalex.org/W4306904969","https://openalex.org/W3019776739","https://openalex.org/W2138720691","https://openalex.org/W2546638913","https://openalex.org/W4362501864","https://openalex.org/W4380318855","https://openalex.org/W3084456289","https://openalex.org/W2209816623","https://openalex.org/W2612731793"],"abstract_inverted_index":{"Components":[0],"of":[1,31,42,62,126,155],"cyber":[2],"physical":[3],"systems,":[4],"which":[5],"affect":[6],"real-world":[7],"processes,":[8],"are":[9,57,72,98,214],"often":[10],"exposed":[11],"to":[12,49,59,74,91,95,101,138,183,208],"the":[13,40,47,84,131,135,153,167,225],"internet.":[14],"Replacing":[15],"conventional":[16],"control":[17,113],"methods":[18,221],"with":[19,39,119,142,170],"Deep":[20],"Reinforcement":[21],"Learning":[22],"(DRL)":[23],"in":[24,159],"energy":[25,44,163],"systems":[26,35],"is":[27],"an":[28,139,147,202],"active":[29],"area":[30],"research,":[32],"as":[33],"these":[34],"become":[36],"increasingly":[37],"complex":[38],"advent":[41],"renewable":[43],"sources":[45],"and":[46,128,174,177,180,200,218],"desire":[48],"improve":[50],"their":[51,63],"efficiency.":[52],"Artificial":[53],"Neural":[54],"Networks":[55],"(ANN)":[56],"vulnerable":[58,100],"specific":[60],"perturbations":[61,71,204],"inputs":[64],"or":[65],"features,":[66],"called":[67],"adversarial":[68,102,191],"examples.":[69,103],"These":[70],"difficult":[73,207],"detect":[75],"when":[76],"properly":[77],"regularized,":[78],"but":[79],"have":[80,194],"significant":[81,195],"effects":[82],"on":[83,197],"ANN's":[85],"output.":[86],"Because":[87],"DRL":[88,172,198,212],"uses":[89],"ANN":[90],"map":[92],"optimal":[93],"actions":[94],"observations,":[96],"they":[97],"similarly":[99],"This":[104],"work":[105],"proposes":[106],"a":[107,120,160],"novel":[108],"attack":[109,132],"technique":[110],"for":[111],"continuous":[112],"using":[114],"Group":[115],"Difference":[116],"Logits":[117],"loss":[118],"bifurcation":[121],"layer.":[122],"By":[123],"combining":[124],"aspects":[125],"targeted":[127,149],"untargeted":[129,140],"attacks,":[130],"significantly":[133],"increases":[134],"impact":[136],"compared":[137],"attack,":[141],"drastically":[143],"smaller":[144],"distortions":[145],"than":[146],"optimally":[148],"attack.":[150],"We":[151],"demonstrate":[152],"impacts":[154,168,196],"powerful":[156],"gradient-based":[157],"attacks":[158,192],"realistic":[161],"smart":[162],"environment,":[164],"show":[165,189],"how":[166],"change":[169],"different":[171],"agents":[173],"training":[175,220],"procedures,":[176],"use":[178],"statistical":[179],"time-series":[181],"analysis":[182],"evaluate":[184],"attacks'":[185],"stealth.":[186],"The":[187],"results":[188],"that":[190],"can":[193,222],"controllers,":[199],"constraining":[201],"attack's":[203],"makes":[205],"it":[206],"detect.":[209],"However,":[210],"certain":[211],"architectures":[213],"far":[215],"more":[216],"robust,":[217],"robust":[219],"further":[223],"reduce":[224],"impact.":[226]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}
