Feature Transformation

The method of PDF projection (without the MaxEnt constraint) has existed since 2000 [3,4]. We consider an arbitrary feature transformation that maps from $ {\cal R}^N$ to $ {\cal R}^D$, where $ D \leq N$ :

$\displaystyle {\bf z}=T({\bf x}), \;\;\; {\bf x}\in{\cal R}^N, \;\;\; {\bf z}\in{\cal R}^D.$ (2.1)



Baggenstoss 2017-05-19