Index

A

AvgMode.Mean (built-in class) AvgMode.None (built-in class) AvgMode.Sum (built-in class)

AvgMode.WeightedMean (built-in class) AvgMode.WeightedSum (built-in class)

C

c (PeriodicLoss attribute)

D

deriv!() (built-in function), [1] deriv() (built-in function), [1], [2], [3], [4], [5], [6] deriv2!() (built-in function) deriv2() (built-in function), [1], [2], [3]

deriv2_fun() (built-in function) deriv_fun() (built-in function) DistanceLoss (built-in class) DWDMarginLoss (built-in class)

E

ExpLoss (built-in class)

H

HuberLoss (built-in class)

I

isclasscalibrated() (built-in function) isclipable() (built-in function) isconvex() (built-in function) isdifferentiable() (built-in function) isdistancebased() (built-in function) islipschitzcont() (built-in function) islocallylipschitzcont() (built-in function)

ismarginbased() (built-in function) isnemitski() (built-in function) isstrictlyconvex() (built-in function) isstronglyconvex() (built-in function) issymmetric() (built-in function) istwicedifferentiable() (built-in function)

L

L1DistLoss (built-in class) L1EpsilonInsLoss (built-in class) L1HingeLoss (built-in class) L2DistLoss (built-in class) L2EpsilonInsLoss (built-in class)

L2HingeLoss (built-in class) L2MarginLoss (built-in class) LogitDistLoss (built-in class) LogitMarginLoss (built-in class) LPDistLoss (built-in class)

M

MarginLoss (built-in class)

ModifiedHuberLoss (built-in class)

N

normalize (AvgMode.WeightedMean attribute) (AvgMode.WeightedSum attribute)

P

PerceptronLoss (built-in class)

PeriodicLoss (built-in class)

Q

q (DWDMarginLoss attribute)

QuantileLoss (built-in class)

S

SigmoidLoss (built-in class) SmoothedL1HingeLoss (built-in class)

SupervisedLoss (built-in class)

V

value!() (built-in function), [1] value() (built-in function), [1], [2], [3], [4], [5], [6] value_deriv() (built-in function), [1], [2]

value_deriv_fun() (built-in function) value_fun() (built-in function)

W

weights (AvgMode.WeightedMean attribute) (AvgMode.WeightedSum attribute)

Z

ZeroOneLoss (built-in class)

Symbols

α (HuberLoss attribute) γ (SmoothedL1HingeLoss attribute)

τ (QuantileLoss attribute) ϵ (L1EpsilonInsLoss attribute) (L2EpsilonInsLoss attribute)