Improvements on isolated word recognition using subspace methods

The purpose of this study is to investigate the effects of different forms of between-class scatter matrices on multi-class problems. Two different between-class scatter matrices are defined in Fisher´s linear discriminant analysis (FLDA) and the classification rates better than that of classical FLDA are obtained for TI-digit database. In this study, the criteria that give separate subspaces for each class are also proposed. It is seen that considering only the within-class scatter in the classification gives better results than that of considering both the within- and between-class scatters for TI-digit database.