The guaranteed estimation of the Lipschitz classifier accuracy: Confidence set approach

Abstract This paper introduces an original method for the guaranteed estimation of the Lipschitz classifier accuracy in the case of a large number of classes. The solution was obtained as a finite closed set of alternative hypotheses, which contains an object of classification with probability of not less than the specified value. Thus, the classification is represented by a set of hypothetical classes. In this case, the smaller the cardinality of the discrete set of hypothetical classes is, the higher is the classification accuracy. This problem is relevant in practical biometrics, when the number of analyzed samples amounts to tens of thousands, and many of them are distinguished vaguely in the primary feature space.

[1]  Alexander J. Smola,et al.  Advances in Large Margin Classifiers , 2000 .

[2]  Jeff A. Bilmes,et al.  A gentle tutorial of the em algorithm and its application to parameter estimation for Gaussian mixture and hidden Markov models , 1998 .

[3]  Andreas Christmann,et al.  Support vector machines , 2008, Data Mining and Knowledge Discovery Handbook.

[4]  Kristin P. Bennett,et al.  Duality and Geometry in SVM Classifiers , 2000, ICML.

[5]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[6]  Haizhou Li,et al.  A GMM supervector Kernel with the Bhattacharyya distance for SVM based speaker recognition , 2009, 2009 IEEE International Conference on Acoustics, Speech and Signal Processing.

[7]  Yiming Ying,et al.  Support Vector Machine Soft Margin Classifiers: Error Analysis , 2004, J. Mach. Learn. Res..

[8]  G. Wahba.  Multivariate Function and Operator Estimation, Based on Smoothing Splines and Reproducing Kernels , 1992 .

[9]  Bernhard Schölkopf,et al.  Maximal margin classification for metric spaces , 2005, J. Comput. Syst. Sci..

[10]  Catherine Blake,et al.  UCI Repository of machine learning databases , 1998 .

[11]  T. Kailath.  The Divergence and Bhattacharyya Distance Measures in Signal Selection , 1967 .

[12]  Tony Jebara,et al.  Bhattacharyya Expected Likelihood Kernels , 2003, COLT.

[13]  John H. L. Hansen,et al.  Speaker identification with whispered speech based on modified LFCC parameters and feature mapping , 2009, 2009 IEEE International Conference on Acoustics, Speech and Signal Processing.

[14]  Bianca Zadrozny,et al.  Transforming classifier scores into accurate multiclass probability estimates , 2002, KDD.

[15]  Matthias Hein,et al.  Maximal Margin Classification for Metric Spaces , 2003, COLT.

[16]  R. Kondor,et al.  Bhattacharyya and Expected Likelihood Kernels , 2003 .

[17]  Ells,et al.  ON EMBEDDING UNIFORM AND TOPOLOGICAL SPACES , 2012 .

[18]  Marti A. Hearst.  Intelligent Connections: Battling with GA-Joe. , 1998 .

[19]  Ulrike von Luxburg,et al.  Distance-Based Classification with Lipschitz Functions , 2004, J. Mach. Learn. Res..