feature selection for normalized margin SVM

Code for "Feature and Region Selection for Visual Learning"

This code provides a Matlab implementation for normalized margin SVM with additive kernels, which corresponds to feature selection part in reference [1]. If linear kernel is used in this code, the feature selection method becomes the method in reference [2].
This code was written by Ji Zhao (Email: [e-mail]). The latest version of this code can be found in his homepage:
https://sites.google.com/site/drjizhao/

Reference
[1] Ji Zhao, Liantao Wang, Ricardo Cabral, and Fernando De la Torre. Feature and Region Selection for Visual Learning. ArXiv: 1407.5245, 2014.
[2] Minh Hoai Nguyen, and Fernando De la Torre. Optimal Feature Selection for Support Vector Machines. Pattern Recognition , 43(3), 584 - 591, 2010.

Keywords: normalized margin SVM, feature selection, additive kernels