ABSTRACT
Appendix A: Computation of Slack Variable-Based SVMs ....................... 43 Appendix B: Computation of Linear ε-SVR ................................................. 44 References .......................................................................................................... 45
and regression based on whether the value of the output vector is discrete or continuous. In this sense, SVMs can be divided into two categories: support vector classiŽcation (SVC) machines and support vector regression (SVR) machines. According to this classiŽcation, the basic elements and algorithms of SVC and SVR are Žrst discussed both theoretically and experimentally in detail, respectively. Then two simulated datasets are employed to investigate the predictive performance of SVM. It is illustrated that SVMs can deal well with data of some nonlinearity. The applications of SVM to real-world data can be found in Chapters 5 through 8, respectively.
