Study Of The Support Vector Machines Under Input Uncertainty

This paper introduces the development of Support Vector Machine systematically while summarizing and analyzing the fruits on this field during the past. Based on the study of some researchers, we study several new algorithms and the models about the uncertainty Support Vector Machines Classification and Regression, basing on the theory of the Support Vector Classification and Regression. The whole paper contains four chapters, and it is arranged as follows:In the first chapter, we summarily introduce the development and the current research situation as well as the solving method of the uncertainty Support Vector Machines, and existing some problems at present.In the second chapter, we introduce the theory of the Support Vector machine Classification, especially analyzing the uncertainty Support Vector Machines Classification, and we study that the uncertain data obey the Log-normal distribution, and get the uncertainty classification model of Support Vector Machine, using the standard normal distribution monotonically and reversibly. At the same time, we propose the classification method that dealing with the uncertain information of the models with the uncertain number. This method can effectively improve the accuracy of classification.In the third chapter, we set up the model of the uncertainty Support Vector Regression by introducing noisy variables. Basing on the theory of the Support Vector Regression, we build the models of the linear and the non-linear Support Vector Regression under the single parameter, and introduce the algorithm of the non-linear Support Vector Regression. Then the experiment explains the effectiveness of the uncertainty Support Vector Regression and it can deal with the disturbance of noise effectively.In the last chapter, it is the conclusion and the prospect and suggestions in the future.