Direct Orthogonal Neighborhood Preserving Discriminant Analysis
Several orthogonal feature extraction algorithms based on local preserving projection have recently been proposed. However, these methods don’t address the singularity problem in the high dimensional feature space,which means that the eigen-equation of orthogonal feature extraction algorithms cannot be solved directly. In this paper, we present a new method called Direct Orthogonal Neighborhood Preserving Discriminant Analysis (DONPDA), which is able to extract all the orthogonal discriminant vectors simultaneously in the high-dimensional feature space and does not suffer the singularity problem. Experimental results on ORL database indicate that the proposed DONPDA method achieves higher recognition rate than the ONPDA method and other some existing orthogonal feature extraction algorithms.
Y. R. Lin and Q. Wang, "Direct Orthogonal Neighborhood Preserving Discriminant Analysis", Advanced Materials Research, Vols. 179-180, pp. 1254-1259, 2011