School of Economic and Management Beijing Jiaotong University, Beijing100044;Logistics School Beijing Wuzi University, Beijing100083;Chinese Academy of Sciences, Research Center on Fictitious Economy and Data Science, Beijing100049
Support Vector Machines have been a dominant learning technique for almost ten years, moreover they have been applied to supervised learning problems. To use the support vector method, assume that training data in the optimization problems are known exactly. But in fact, the training data are usually subject to measurement noise. In this paper, a robust semi-supervised classification algorithm based on linear -Support Vector Machines is presented. Numerical simulation shows the robustness of the proposed method.
PDF(460 KB)
