Target Classification by a New Class of Linear Discriminants
A new linear discriminant technique that results in better classification performance over existing techniques is presented in this paper. This new approach is formulated in a similar manner to that of the Fisher linear discriminant. However, the matrix which corresponds to within classes has been replaced by a new matrix. This matrix takes into consideration the cross-correlation properties of the classes of interest. It has been shown through simulations that this matrix replacement results in a better classification performance over other linear discrimination methods including the Fisher discriminant. Finally, the proposed new discriminant is presented in parametric and nonparametric forms, and is found to exhibit better classification in both cases over other parametric and nonparametric methods. With this new approach, the nonparametric method will prove to be more successful than its parametric counterpart. The feature selection is also discussed.
KeywordsFeature Selection Scatter Matrix Good Classification Performance Minimum Classification Error Fisher Linear Discriminant
Unable to display preview. Download preview PDF.
- 1.A. Bhattacharyya, A measure of Divergence Between Two Statistics Populations Defined by Their Probability Distributions, Bull. Calcutta Math Soc., 45, pp.99–110 (1943).Google Scholar
- 3.F. Rosenblatt, The Perceptron: A Perceiving and Recognizing Automation, Project PARA, Cornell Aeronauts Lab., Rept 85–460–1 (1957).Google Scholar
- 4.F. Rosenblatt, On the Convergence of Reinforcement Procedures in Simple Perceptrons, Cornell Aeronauts Lab., Rept. VG-1196-G-4 (1960).Google Scholar
- 5.A. Charres, On Some Fundamental Theorems of Perceptron Theory and Their Geometry, Computer and Information Sciences-I (J. T. Tou and R. Wilcox, eds., Spartan Books, Washington D.C. (1964).Google Scholar
- 8.B. G. Batchelor and B. R. Wilkins, Method for Location Clusters of Pattern to Initialize a Learning Machine, Electronics Letters, Vol. 5, No. 20, pp. 481–483.Google Scholar
- 9.G. H. Ball and D. J. Hall, ISODATA, an Iterative Method of Multivariate Analysis and Pattern Classification, Proceeding of the IFIPS Congress (1965).Google Scholar
- 10.R. C. Gonzales and J. T. Tou, Some Results in Minimum Entropy Feature Extraction, IEEE Convention Record, Region III (1968).Google Scholar
- 15.J.C. Bezdek and S.K. Pal (Eds.), Fuzzy Models For Pattern Recognition, IEEE Press (1992).Google Scholar
- 20.T. Young and K-S Fu, Handbook of Patern Recognition and Image Processing, Academic Press (1986).Google Scholar