{"references": ["B. Scholk\u00f6pf, A. Smola, and K.R. M\u251c\u255dller, \"Nonlinear Component Analysis\nas a Kernel Eigenvalue Problem,\" Neural Computation, vol.10, pp.\n1299-1319, 1998.", "G. Baudat and F. Anouar, \"Generalized Discriminant Analysis Using a\nKernel Approach,\" Neural Computation, vol.12, pp. 2385-2404, 2000.", "C. Park and H. Park, \"Fingerprint Classification Using Nonlinear Discriminant\nAnalysis,\" Technical Report, TR 03-034, University of Minnesota,\nUSA, Sep. 2003.", "S. Mika, G. R\u251c\u00f1tsch, J. Weston, B. Sch\u00f6lkopf, and K.R. M\u251c\u255dller, \"Fisher\ndiscriminant analysis with kernels,\" Neural Networks for Signal Processing\nIX, pp. 41-48, 1999.", "P.N. Belhumeour, J.P. Hespanha, and D.J. Kriegman, \"Eigenfaces vs.\nFisherfaces: Recognition Using Class Specific Linear Projection,\" IEEE\nTrans. Pattern Analysis and Machine Intell., vol.19, pp. 711-720, 1997.", "H.C. Kim, D. Kim, and S.Y. Bang, \"Face recognition using LDA mixture\nmodel,\" Pattern Recognition Letters, vol.24, pp. 2815-2821, 2003.", "G. R\u251c\u00f1tsch, T. Onoda, and K.R. M\u251c\u255dller, \"Soft Margins for AdaBoost,\"\nMachine Learning, pp. 1-35, 2000.", "C. Burges, \"Simplified Support Vector Decision Rules,\" in Proceedings\nof the 13th International Conference on Machine Learning, pp. 71-77,\n1996.", "B. Sch\u00f6lkopf, P. Knirsch, A. Smola, and C. Burges, \"Fast Approximation\nof Support Vector Kernel Expansions, and an Interpretation of\nClustering as Approximation in Feature Spaces,\" Proceedings of the\nDAGM Symposium Mustererkennung, pp. 124-132, 1998."]}

The Linear discriminant analysis (LDA) can be generalized into a nonlinear form - kernel LDA (KLDA) expediently by using the kernel functions. But KLDA is often referred to a general eigenvalue problem in singular case. To avoid this complication, this paper proposes an iterative algorithm for the two-class KLDA. The proposed KLDA is used as a nonlinear discriminant classifier, and the experiments show that it has a comparable performance with SVM.