Study Of The Unascertained Optimization Problem Based On The Unascertained Theory

Unascertained theory is proposed by Professor Wang Guangyuan of China Academy of Engineering in 1990, which is used to characterize the subjective awareness of policy makers on dealing with the uncertainty arising from the uncertainty information. "Unascertained information and mathematical treatment" opened the first page of the unascertained mathematical research,which is proposed by Professor Wang Guangyuan in 1990, caused great concern in a number of domestic and foreign academic and engineering sector.The unascertained mathematical theory has been initially established, and has a preliminary application in many fields, and achieved certain results.The paper introduces the unascertained theory's proposed and development, introduces the concepts of unascertained chance constrained programming, which is porposed by Professor Yang Zhimin, and detaily introduces the model of the unascertained chance constrained programming and it's solutions. Based on those study, we propose a methods on how to solve the unascertained chance constrained programming when it's constraind has a more general form. Numerical example is given shows that the methods is effective.Then the article discussed several ways on how to deay with the support vector machine when it has unascertained infromation. First of all, the article introduce the pretreatment method on unascertained information. Process the unascertained sample in the training set, making it into a training set for class labels determine, then use the traditional model of support vector machine. Then based on the unascertained support vector machine which is proposed by Professor Yang Zhimin, use the methods of reconstructed variable, give a algorithm of support vector machine, at the same time provides a numerical example and do some analysis on the example results, describes the feasibility and effectiveness of the algorithm.Finally, the paper study the classification when the point has a unascertaine class, the article based on unascertained theory, raised the issue of uncertain calssification on an unascertained of chance constrained programming model, and with the help of unascertained raised earlier algorithm of support vector machine, solved the unascertained for solving support vector machines for calssification problems.