OvercomingCatastrophicForgettingbyBayesianGenerativeRegularizationPatrickH.Chen1WeiWei2Cho-jui,Hsieh1BoDai3AbstractelsfailtotacklecontinuallearningwiththephenomenonofcatastrophicForgetting,wheredee...
AddressingCatastrophicForgettinginFew-ShotProblemsPauchingYap1HippolytRitter1DavidBarber12Abstractclassesiscalledthemeta-trainingstage,andtheevaluationprocessthatreportsthefew-shotperformanceonthen...
SIGUA:ForgettingMayMakeLearningwithNoisyLabelsMoreRobustBoHan12GangNiu2XingruiYu3QuanmingYao4MiaoXu25IvorW.Tsang3MasashiSugiyama26Abstractasweightdecay(Krogh&Hertz,1991)anddropout(Sri-vastavaetal.,...
OvercomingMulti-modelForgettingYassineBenyahia†1KaichengYu2KamilBennani-Smires3MartinJaggi4AnthonyDavison1MathieuSalzmann2ClaudiuMusat3Abstractposteriorprobability,p(θD1,D2),withθthenetworkparam...
TheHierarchicalAdaptiveForgettingVariationalFilterVincentMoens1Abstractfluctuationswhentheyhappenfasterthanexpected.Mostmodelscoveroneofthetwopossibilities,andeithergradu-AcommonprobleminMachineLea...
OvercomingCatastrophicForgettingwithHardAttentiontotheTaskJoanSerra`1D´ıdacSur´ıs12MariusMiron13AlexandrosKaratzoglou1Abstractintheadvancementtowardsmoregeneralartificialintel-ligencesystems(Le...