Quasi-GlobalMomentum:AcceleratingDecentralizedDeepLearningonHeterogeneousDataTaoLin1SaiPraneethKarimireddy1SebastianU.Stich1MartinJaggi1Abstractiskeptlocally(nevertransmittedduringtraining).Decentr...
QuantumAlgorithmsforReinforcementLearningwithaGenerativeModelDaochenWang1AarthiSundaram2RobinKothari2AshishKapoor3MartinRoetteler2Abstractfasteralgorithmsforcertaintaskslikesearchandfactor-ing(Grov...
QuantifyingtheBenefitofUsingDifferentiableLearningoverTangentKernelsEranMalach1PritishKamath2EmmanuelAbbe3NathanSrebro2AbstractCanallthesuccessofdeepLearningbeexplainedusingtheNTK?Thiswouldimplytha...
PsiPhi-Learning:ReinforcementLearningwithDemonstrationsusingSuccessorFeaturesandInverseTemporalDifferenceLearningAngelosFilos1ClareLyle1YarinGal1SergeyLevine2NatashaJaques23GregoryFarquhar4Abstract...
Puttingthe“Learning”intoLearning-AugmentedAlgorithmsforFrequencyEstimationElbertDu12FranklynWang12MichaelMitzenmacher2AbstractbyθandanalgorithmAInLearning-augmentedalgorithms,algorithmsmaxM(θ,A...
ProximalCausalLearningwithKernels:Two-StageEstimationandMomentRestrictionAfsanehMastouri1YuchenZhu1LimorGultchin23AnnaKorba4RicardoSilva1MattJ.Kusner1ArthurGretton†1KrikamolMuandet†5AbstractGoalA...
ProvablyEnd-to-endLabel-noiseLearningwithoutAnchorPointsXuefengLi12TongliangLiu2BoHan3GangNiu4MasashiSugiyama45Abstract(Arpitetal.,2017;Zhangetal.,2017;Xiaetal.,2021;Wuetal.,2021).Inlabel-noiselear...
ProvablyEfficientReinforcementLearningforDiscountedMDPswithFeatureMappingDongruoZhou1JiafanHe1QuanquanGu1Abstractlinearfunctionsorneuralnetworkstomapstatesandactionstoalow-dimensionalspaceandsolvet...
ProvablyEfficientLearningofTransferableRewardsAlbertoMariaMetelli1GiorgiaRamponi1AlessandroConcetti1MarcelloRestelli1Abstracttheoretically,underthestrongassumptionofrewardunique-ness(Abbeel&Ng,2004...
ProvableRobustnessofAdversarialTrainingforLearningHalfspaceswithNoiseDifanZou1SpencerFrei2QuanquanGu1AbstractToformalizetheabovecomment,letusdefinethero-Weanalyzethepropertiesofadversarialtrain-bus...
PreferentialTemporalDifferenceLearningNishanthAnand12DoinaPrecup123AbstractTD-Learningcanbeviewedasawaytoapproximatedy-namicprogrammingalgorithmsinMarkovianenviron-Temporal-Difference(TD)Learningis...
PracticalandPrivate(Deep)LearningWithoutSamplingorShufflingPeterKairouz1BrendanMcMahan1ShuangSong1OmThakkar1AbhradeepThakurta1ZhengXu1Abstractinthecontextofdistributedsettingslikefederatedlearn-ing...
Prediction-CentricLearningofIndependentCascadeDynamicsfromPartialObservationsMateuszWilinski1AndreyY.Lokhov1AbstractThismotivatesthedevelopmentofefficientalgorithmsthatcaninfertheeffectivespreading...
PersonalizedFederatedLearningusingHypernetworksAvivShamsian∗1AvivNavon∗1EthanFetaya1GalChechik12Abstractwhileminimizingcommunication.Unfortunately,Learningasingleglobalmodelmayfailwhenthedatadist...
PEBBLE:Feedback-EfficientInteractiveReinforcementLearningviaRelabelingExperienceandUnsupervisedPre-trainingKiminLee1LauraSmith1PieterAbbeel1AbstractKober&Peters,2011;Koberetal.,2013;Silveretal.,201...
PC-MLP:Model-basedReinforcementLearningwithPolicyCoverGuidedExplorationYudaSong1WenSun2Abstractsuccessrate0.5HandEgg0.4Model-basedReinforcementLearning(RL)isa0.3DeepPC-MPL200000popularLearningparad...
ParameterlessTransductiveFeatureRe-representationforFew-ShotLearningWentaoCui1YuhongGuo12AbstractFSLsolutionsaregenerallydevelopedintwobranches:meta-Learningbasedmethodsandfinetuningbasedmeth-Recen...
On-PolicyDeepReinforcementLearningfortheAverage-RewardCriterionYimingZhang1KeithW.Ross21AbstractHaarnojaetal.,2018)orinaqueuingscenario(Tadepalli&Ok,1994;Sutton&Barto,2018),thereisnonaturalsep-Wede...
OnlinePolicyGradientforModelFre√eLearningofLinearQuadraticRegulatorswithTRegretAsafCassel1TomerKoren12AbstractModel-basedmethods,whichperformplanningbasedonasystemidentificationprocedurethatestima...
OnlineLearningforLoadBalancingofUnknownMonotoneResourceAllocationGamesAbstractIlaiBistritz1NicholasBambos12018;Boursier&Perchet,2019;Mehrabianetal.,2020;ConsiderNplayersthateachusesamixtureofNayyar...