trait
OptimizableObjective[Prediction, Output] extends AnyRef
Abstract Value Members
-
abstract
def
valueAndGradient(prediction: Prediction, label: Output): (Double, Prediction)
Concrete Value Members
-
final
def
!=(arg0: AnyRef): Boolean
-
final
def
!=(arg0: Any): Boolean
-
final
def
##(): Int
-
final
def
==(arg0: AnyRef): Boolean
-
final
def
==(arg0: Any): Boolean
-
final
def
asInstanceOf[T0]: T0
-
def
clone(): AnyRef
-
final
def
eq(arg0: AnyRef): Boolean
-
def
equals(arg0: Any): Boolean
-
def
finalize(): Unit
-
final
def
getClass(): Class[_]
-
def
hashCode(): Int
-
final
def
isInstanceOf[T0]: Boolean
-
final
def
ne(arg0: AnyRef): Boolean
-
final
def
notify(): Unit
-
final
def
notifyAll(): Unit
-
final
def
synchronized[T0](arg0: ⇒ T0): T0
-
def
toString(): String
-
final
def
wait(): Unit
-
final
def
wait(arg0: Long, arg1: Int): Unit
-
final
def
wait(arg0: Long): Unit
Abstract trait for any (sub)differentiable objective function used to train predictors. More accurately, this defines a family of objective functions indexed by each possible label. Note that these are concave objective functions, not convex loss functions.
The type of the prediction: is it a tensor or a scalar?
The type of the output/label: is it integer or real-valued or tensor-valued?