Meta-LearningBidirectionalUpdateRulesMarkSandler1MaxVladymyrov1AndreyZhmoginov1NolanMiller1AndrewJackson1TomMadams1BlaiseAgu¨erayArcas1AbstracttracedbacktoRobbins&Monro(1951).Sinceitsintro-duction...
SafeScreeningRulesfor0-RegressionfromPerspectiveRelaxationsAlperAtamtu¨rk1Andre´sGo´mez2Abstract2015),andthe2(ridge)regularization(Hoerl&Kennard,1970)imposesbias/shrinkageintheregressioncoeffici...
ScreeningRulesforLassowithNon-ConvexSparseRegularizersAlainRakotomamonjy12GillesGasso1JosephSalmon3AbstractWhiletheLassohashadagreatimpactonmachinelearn-ingandsignalprocessingcommunitiesgivensomesu...
DecouplingGradient-LikeLearningRulesfromRepresentationsPhilipS.Thomas1ChristophDann2EmmaBrunskill3Abstractworkpoorlywith,orbeincompatiblewith,others.Althoughthisintertwiningisunavoidableandislikely...