Based On Support Vector Machine Regression

With the in-depth study on the theory of machine learning (LM) and the constantimprovement on Data Mining (DM) technology, Vapik.V puts forward the SupportVector Machine (SVM) technology in the early 90s. Based on the statistical learningtheory (SLT), it is a new method to solve the problems of machine learning and the newdata mining with the help of optimization. Different from traditional statistics andartificial neural network (ANN), Support Vector Machine is to study the statisticalmethod and law of the small samples. Furthermore, the sample used in Support VectorMachine is not the same as in that used in statistics, which is specifically designed for acertain target. It is proved that the support vector machine based on the statisticallearning theory is not only simple in structure, but has a strong ability to promote thetheory and put in practice. At present, the research on Support Vector Machine hasbecome a hot issue in the field of international artificial intelligence and database. Inthis paper, the Support Vector Machine technology is explored systematically anddeeply.The main thesis of this paper is as follows: First, there is a brief introduction tothe basic concepts of the machine learning and data mining as well as their internationalcurrent research. The support vector machines technique is generally brought forwardIn the second chapter, there is an introduction to the regression issues in thestatistics, such as linear regression and logistic regression models.In the third part, through uniting the support vector machine technology with theregression analysis issues, the data regression in System Support Vector Machine issystematically discussed; meanwhile, optimization tools are adopted to simplify theissue to get a simple but reasonable model.Finally, in order to resolve non-smoothness in the Support Vector Machine datareturn, especially the non-differentiable problem of some inflection points, theapplication of the smooth technology in Support Vector Machine regression is in-depthstudied in chapter four, and three smooth support vector machine regression methodswhich are efficient in theory and practice are brought forward. At present, the research in the support vector machine is mainly in the field ofstatistics, mental intelligence, and statistical data. However, there is little study on thesmoothness of the support vector machine regression issue. This paper has amagnificent meaning in the research on the support vector machine technology in theoryas well as in practice.