GClasses
|
A loss function used to train a differentiable function.
#include <GOptimizer.h>
Public Member Functions | |
GObjective () | |
virtual | ~GObjective () |
virtual void | calculateOutputLayerBlame (const GVec &prediction, const GVec &label, GVec &blame)=0 |
Calculate the error term (a.k.a. blame) associated with the activation of this layer. More... | |
virtual void | evaluate (const GVec &prediction, const GVec &label, GVec &loss)=0 |
Calculate the error. More... | |
virtual void | setSlack (const GVec &slack) |
Enable the use of slack (a margin-of-error). More... | |
Protected Attributes | |
GVec | m_slack |
bool | m_useSlack |
|
inline |
|
inlinevirtual |
|
pure virtual |
Calculate the error term (a.k.a. blame) associated with the activation of this layer.
Implemented in GClasses::GSquaredError.
|
pure virtual |
Calculate the error.
Implemented in GClasses::GSquaredError.
|
inlinevirtual |
Enable the use of slack (a margin-of-error).
|
protected |
|
protected |