libeblearn
|
#include <ebl_nonlinearity.h>
Public Member Functions | |
smooth_shrink_module (parameter< T, Tstate > *p, intg nf, T beta=10, T bias=.3) | |
virtual | ~smooth_shrink_module () |
Destructor. | |
virtual void | fprop (Tstate &in, Tstate &out) |
forward | |
virtual void | bprop (Tstate &in, Tstate &out) |
backward | |
virtual void | bbprop (Tstate &in, Tstate &out) |
2nd deriv backward | |
virtual smooth_shrink_module < T, Tstate > * | copy () |
Returns a deep copy of this module. | |
Public Attributes | |
Tstate | beta |
Tstate | bias |
A smoothed shrinkage module that parametrizes the steepnes and location of the shrinkage operator. This function is useful for learning since there is always gradients flowing through it.
ebl::smooth_shrink_module< T, Tstate >::smooth_shrink_module | ( | parameter< T, Tstate > * | p, |
intg | nf, | ||
T | beta = 10 , |
||
T | bias = .3 |
||
) |
Constructor.
nf | The number of features. |