Title :
Kernelized Supervised Dictionary Learning
Author :
Gangeh, M.J. ; Ghodsi, Ali ; Kamel, Mohamed S.
Author_Institution :
Dept. of Med. Biophys., Univ. of Toronto, Toronto, ON, Canada
Abstract :
In this paper, we propose supervised dictionary learning (SDL) by incorporating information on class labels into the learning of the dictionary. To this end, we propose to learn the dictionary in a space where the dependency between the signals and their corresponding labels is maximized. To maximize this dependency, the recently introduced Hilbert Schmidt independence criterion (HSIC) is used. One of the main advantages of this novel approach for SDL is that it can be easily kernelized by incorporating a kernel, particularly a data-dependent kernel such as normalized compression distance, into the formulation. The learned dictionary is compact and the proposed approach is fast. We show that it outperforms other unsupervised and supervised dictionary learning approaches in the literature, using real-world data.
Keywords :
dictionaries; face recognition; learning (artificial intelligence); pattern recognition; HSIC; Hilbert Schmidt independence criterion; SDL; data-dependent kernel; face recognition; kernelized supervised dictionary learning; normalized compression distance; real-world data; Dictionaries; Educational institutions; Electronic mail; Image reconstruction; Kernel; Matching pursuit algorithms; Training; Classification methods; HSIC; dictionary learning; non-parametric methods; pattern recognition and classification; supervised learning;
Journal_Title :
Signal Processing, IEEE Transactions on
DOI :
10.1109/TSP.2013.2274276