Regularized Discriminant Analysis for Face Recognition
Itzik Pima , Mayer Aladjem
Department of Electrical and Computer Engineering, Ben-Gurion University of the Negev
P.O.Box 653, Beer-Sheva, 84105, Israel.
This paper studies Regularized Discriminant Analysis (RDA) in the context of face
recognition. We check RDA sensitivity to different photometric preprocessing methods and
compare its performance to other classifiers. Our study shows that RDA is better able to
extract the relevant discriminatory information from training data than the other classifiers
tested, thus obtaining a lower error rate. Moreover, RDA is robust under various lighting
conditions while the other classifiers perform badly when no photometric method is applied.
Keywords: face recognition, feature extraction, regularization, principal component
analysis, discriminant analysis, photometric preprocessing.
This study compares the performance of Regularized Discriminant Analysis 
(RDA) with that of two classifiers: L2 (Euclidean distance) and angle (Normalized
Correlation), usually used for face recognition. In saying L2 and angle we mean that
we use a nearest center classifier using those distance metrics. The potential of
extracting the relevant discriminatory information from a small amount of training