Diagonal principal component analysis with non-greedy l(1)-norm maximization for face recognition

摘要

Diagonal principal component analysis (DiaPCA) is an important method for dimensionality reduction and feature extraction. It usually makes use of the l(2)-norm criterion for optimization, and is thus sensitive to outliers. In this paper, we present a DiaPCA with non-greedy l(1)-norm maximization (D1aPCA-L1 non-greedy), which is more robust to outliers. Experimental results on two benchmark datasets show the effectiveness and advantages of our proposed method. (C) 015 Elsevier B.V. All rights reserved.