A novel competitive learning algorithm for the parametric classification with Gaussian distributions

Abstract A competitive learning algorithm for the parametric classification of Gaussian sources is presented in this letter. The algorithm iteratively estimates the mean and prior probability of each class during the training. Bayes rule is then used for classification based on the estimated information. Simulation results show that the proposed algorithm outperforms k -means and LVQ algorithms for the parametric classification.

[1]  Les E. Atlas,et al.  A comparison of the LBG algorithm and Kohonen neural network paradigm for image vector quantization , 1990, International Conference on Acoustics, Speech, and Signal Processing.

[2]  Manolis A. Christodoulou,et al.  Convergence properties of a class of learning vector quantization algorithms , 1996, IEEE Trans. Image Process..

[3]  Richard O. Duda,et al.  Pattern classification and scene analysis , 1974, A Wiley-Interscience publication.

[4]  Teuvo Kohonen,et al.  The self-organizing map , 1990 .

[5]  Chin-Teng Lin,et al.  Neural fuzzy systems , 1994 .

[6]  Simon Haykin,et al.  Neural network approaches to image compression , 1995, Proc. IEEE.

[7]  Wen-Jyi Hwang,et al.  A novel entropy-constrained competitive learning algorithm for vector quantization , 1999, Neurocomputing.

[8]  Allen Gersho,et al.  Vector quantization and signal compression , 1991, The Kluwer international series in engineering and computer science.

[9]  Teuvo Kohonen,et al.  Improved versions of learning vector quantization , 1990, 1990 IJCNN International Joint Conference on Neural Networks.

[10]  Robert J. Schalkoff,et al.  Pattern recognition - statistical, structural and neural approaches , 1991 .