The Applications Of Support Vector Machines For Regression In Curves Fitting/Reconstruction

Support Vector Machine (SVM for short) is a machine learning method based on Statistical Learning Theory. Support Vector Machine for Regression (SVR for short) is the regression model of Support Vector Machine. As these methods are based on strict theory and successful on solving many actual problems, they have become one of the most important achievements in machine study field in the last ten years.In this paper, the SVR is the smoothness fitting model in the Computer Aided Geometry Design (CAGD for short).According to the features and interface of B splines in CAGD, we construct a new B splines Kernel. In the meantime, we improve the "punish coefficient C" of the SVR to make it much better applied in smoothness fitting/reconstruction.