One New Method for the Rotation in Discriminant Analysis
Keywords:
canonical discriminant analysis, centroid rotationAbstract
Different methods of discriminant weights and structure matrices rotation has been applied in canonical discriminant analysis to facilatate the interpretation of the discriminant functions. A new ("Centrovax") method of discriminant structure and discriminant weights matrices rotation has been proposed in this paper. The proposed method is based on the rotation of group centroid matrix based on the significant discriminant functions. A program implementation of proposed algorithm has been written in SPSS MATRIX language. An illustrative numerical example of use of the method has been presented.
References
Anderson, T. W. (1984). An introduction to multivariate statistical analysis. Wiley.
Lunneborg, C. E., & Lunneborg, P. W. (1978). Improved counseling information through rotation of discriminant functions. Educational and Psychological Measurement, 38(3), 737–754.
Mulaik, S. A. (1972). The foundations of factor analysis. McGraw-Hill.
Momirović, K. (1997). O diskriminativnim funkcijama, diskriminativnim faktorima i nekim očiglednim glupostima. Statistička revija, 46(1–2), 79–100.
Momirović, K., & Zorić, A. (1996). On the variance, reliability, significance and importance of canonical discriminant functions. U S. Bogosavljević & M. Kovačević (Ur.), Analiza grupisanja (Vol. 2, str. 79–91). Savezni zavod za statistiku.
Rao, C. R. (1974). Linear statistical inference and its applications (2. izd.). Wiley Eastern Private Limited.
Rencher, A. C. (1992). Interpretation of canonical discriminant functions, canonical variates, and principal components. The American Statistician, 46(3), 217–225.
Tatsuoka, M. M. (1971). Multivariate analysis: Techniques for educational and psychological research. John Wiley & Sons.
Tenjović, L. (2002). Statistika u psihologiji – priručnik. Centar za primenjenu psihologiju.


