2023
Multi-Agent Reinforcement Learning with Epistemic Priors
Walker T, Ide J, Choi M, Guarino M, Alcedo K. Multi-Agent Reinforcement Learning with Epistemic Priors. 2023, 00: 2514-2518. DOI: 10.1109/codit58514.2023.10284342.Peer-Reviewed Original Research
2022
Hierarchical Reinforcement Learning for Air Combat at DARPA's AlphaDogfight Trials
Pope A, Ide J, Miovi D, Diaz H, Twedt J, Alcedo K, Walker T, Rosenbluth D, Ritholtz L, Javorsek D. Hierarchical Reinforcement Learning for Air Combat at DARPA's AlphaDogfight Trials. IEEE Transactions On Artificial Intelligence 2022, 4: 1371-1385. DOI: 10.1109/tai.2022.3222143.Peer-Reviewed Original ResearchAir combatLow-level policiesHierarchical reinforcement learningField of roboticsAdoption of AIHierarchical Deep ReinforcementContinuous control problemsDeep reinforcementArtificial intelligenceReward shapingContinuous state spaceReinforcement learningState spaceAutonomous controlExpert knowledgeCombat systemExpert pilotsAIImportant challengeControl problemRoboticsIntelligenceLearningComplexitySpaceSoft Actor-Critic with Inhibitory Networks for Retraining UAV Controllers Faster
Choi M, Filter M, Alcedo K, Walker T, Rosenbluth D, Ide J. Soft Actor-Critic with Inhibitory Networks for Retraining UAV Controllers Faster. 2022, 00: 1561-1570. DOI: 10.1109/icuas54217.2022.9836052.Peer-Reviewed Original ResearchSoft Actor-CriticDeep reinforcement learningUnmanned aerial vehiclesActor-CriticAutonomous unmanned aerial vehiclesRealistic simulation environmentLow-level controlNon-stationary environmentsDRL agentDRL algorithmCatastrophic forgettingReinforcement learningBaseline methodsSimulation environmentUAV controllerSample efficiencyAerial vehiclesReal worldDifficult taskProportional-IntegralNovel approachNetworkActive researchValue evaluationAlgorithm
2021
Hierarchical Reinforcement Learning for Air-to-Air Combat
Pope A, Ide J, Mićović D, Diaz H, Rosenbluth D, Ritholtz L, Twedt J, Walker T, Alcedo K, Javorsek D. Hierarchical Reinforcement Learning for Air-to-Air Combat. 2021, 00: 275-284. DOI: 10.1109/icuas51884.2021.9476700.Peer-Reviewed Original Research
2018
Oxytocin attenuates trust as a subset of more general reinforcement learning, with altered reward circuit functional connectivity in males
Ide J, Nedic S, Wong K, Strey S, Lawson E, Dickerson B, Wald L, La Camera G, Mujica-Parodi L. Oxytocin attenuates trust as a subset of more general reinforcement learning, with altered reward circuit functional connectivity in males. NeuroImage 2018, 174: 35-43. PMID: 29486321, DOI: 10.1016/j.neuroimage.2018.02.035.Peer-Reviewed Original ResearchConceptsUltra-high field fMRIPre-existing beliefsReinforcement learningBrain encodingGeneral reinforcement learningNeuroeconomic taskIntranasal oxytocinFace of informationField fMRIFeedback learningFMRI activationExperimental paradigmNegative rewardsOrbitofrontal cortexSocial learningReward circuitBehavioral trajectoriesFunctional connectivityConnectivity analysisSocial contextSubjects' behaviorSocial relationshipsBehavioral effectsLearningBayesian expectation