Kernel classifier with Correntropy loss

Rosha Pokharel, Jose C. Principe
2012 The 2012 International Joint Conference on Neural Networks (IJCNN)  
Classification can be seen as a mapping problem where some function of xn predicts the expectation of a class variable yn. This paper uses kernel methods for the prediction of class variable, together with a recently proposed cost function for classification, called Correntropy-loss(C-loss) function. C-Loss is a non-convex loss function based on a similarity measure called correntropy and is known to closely approximate the ideal 0 − 1 loss function for classification. This paper shows via
more » ... imental results that, by replacing the cost function -Mean Square Error (MSE) in a conventional kernel based functional mapping, by a non-convex loss function C-Loss, a non-overfitting, and hence, a better classifier can be obtained. Since gradient descent can still be used with the C-loss and the kernel mapper, the classifier can be easily trained without performance penalty, compared to the SVM, which makes the approach very practical.
doi:10.1109/ijcnn.2012.6252721 dblp:conf/ijcnn/PokharelP12 fatcat:4etjnodwsncytcnsx77ishdoy4