Efficient linear discriminant analysis with locality preserving for face recognition

摘要: Linear discriminant analysis (LDA) is one of the most popular techniques for extracting features in face recognition. LDA captures global geometric structure. However, local structure has recently been shown to be effective In this paper, we propose a novel feature extraction algorithm which integrates both and structures. We first cast as least square problem based on spectral regression, then regularization technique used model Furthermore, impose penalty parameters tackle singularity design an efficient selection choose optimal tuning parameter balances tradeoff between Experimental results four well-known data sets show that proposed integration framework competitive with traditional recognition algorithms, use either or only.