Study On Multi-class Pattern Statistical Recognition Model And Application

In realistic life, the problem of pattern recognition, especially the study of multi-classification pattern recognition is very significant. The field of pattern recognition is very comprehensive, for example, disease diagnosis, physical check-up, fingerprint recognition and the problem of fault diagnosis and detect prediction in the power system and traffic system etc., all that above can be regarded as the study of pattern recognition. The SVM method based on the statistical learning theory is one classification method that own the good generalization ability. It displays the unique superiority and the good application prospect in solving the small sample, non-linear and in the high dimension pattern recognition question. But the basic support vector machine is proposes based on two class problems, the theory and application research of multi-class classification on SVM is at the exploratory stage because of noise data, uncertain classification and skewed training sets in fault classification. Therefore, it is a very significative task for us to have a research of multi-class classification.Aimed at the problems that encountered in the statistical pattern recognition, such as the difficulty of obtaining samples, noise data, and skewed training sets in fault classification etc.. It combined the ideal of the fuzzy support vector machine and support vector data description (SVDD) and proposed a fuzzy weighted SVDD muti-class classification method. The contents and the results of the paper mainly contained:(1) The paper has carried on the thorough analysis to the domestic and foreign research actuality of the multi-class statistical pattern recognition methods, especially carried on the thorough study to the SVM theory and algorithm. And carried on the thorough analysis to the SVM multi-classification methods, and compared their advantages and disadvantages. Introduced the ideal and theory of FSVM to built the foundation for the further research.(2) Support vector data description (SVDD) was originally designed for one-class classification and novelty detection. Through analysis for its characteristic. SVDD is extended to multi-class classification problems. fuzzy value (weight) is computed by an improved noise clustering algorithm aiming at noise data of training sets. Then a classification decision formula is constructed and a fuzzy weighted SVDD multi-class classification method. Bayesian theories analysis indicates the proposed classification decision-making formula satisfies Bayesian decision- making rules.(3)The paper made use of the data sets in the UCI database and fault diagnosis data to carry on analyzing and appraising in the recognition effect to the model.