Abstract Many linear discriminant analysis (LDA) and kernel Fisher discriminant analysis (KFD) methods are based on the restrictive assumption that the data are homoscedastic. In this paper, we propose a new KFD method called heteroscedastic kernel weighted discriminant analysis (HKWDA) which has several appealing characteristics. First, like all kernel methods, it can handle nonlinearity efficiently in a disciplined manner. Second, by incorporating a weighting function that can capture heteroscedastic data distributions into the discriminant criterion, it can work under more realistic situations and hence can further enhance the classification accuracy in many real-world applications. Moreover, it can effectively deal with the small sample size problem. We have performed some face recognition experiments to compare HKWDA with several linear and nonlinear dimensionality reduction methods, showing that HKWDA consistently gives the best results.
Extending Kernel Fisher Discriminant Analysis with the Weighted Pairwise Chernoff Criterion
Computer Vision – ECCV 2006 ; 6 ; 308-320
Lecture Notes in Computer Science ; 3954 , 6
2006-01-01
13 pages
Article/Chapter (Book)
Electronic Resource
English
Extending Kernel Fisher Discriminant Analysis with the Weighted Pairwise Chernoff Criterion
British Library Conference Proceedings | 2006
|Fisher+Kernel criterion for discriminant analysis
IEEE | 2005
|Ship target recognition using kernel Fisher discriminant [4554-26]
British Library Conference Proceedings | 2001
|Image Interpolation using Texture Orientation Map and Kernel Fisher Discriminant
British Library Conference Proceedings | 2005
|