a study on the effect of between-class scatter matrix on two-class lda method

Biometrics has become attractive in areas that require high security and control. of all the technologies out there, facial detection is one of the most widely used and consistent technologies. linear discriminant analysis (lda) performs dimension reduction while maintaining class discriminatory information as far as possible. in this work, we propose the use of projections of training samples into a subspace defined by the between-class scatter data and classification based on support vector machine (svm) for two classes. our experimental work on four well-known face databases clearly shows the effect of between-class matrix in two-class classification.