A supervised learning framework for PCA-based face recognition using GNP fuzzy data mining

Traditional PCA-based face recognition algorithms usually have low performance in the complicated illumination database. There are two reasons. One is that the number of classes is large compared with other classification problems. The other is that the data in the PCA domain distributes in a narrow space and overlaps frequently. This paper presents a novel supervised learning framework for PCA-based face recognition using Genetic Network Programming (GNP) fuzzy data mining (GNP-FDM). In the proposed framework, a face recognition oriented genetic-based clustering algorithm (GCA) is used to reduce the number of classes and overlaps in the recognition. And, a fuzzy class association rules (FCARs) based classifier is applied to mine the inherent relationships between eigen-vectors and to improve the recognition accuracy. Experimental results on the extended Yale-B database indicate that the proposed supervised learning framework has higher accuracy compared with the traditional PCA-based methods.