Integration of CMAC-GBF Systems and Support Vector Regression Approach

In this study, we integrate the techniques of the cerebellar model articulation controller with general basis function (CMAC-GBF) systems and the support vector regression (SVR) approach to be a more efficient scheme. The advantages of the CMAC-GBF systems include: fast learning speed, guarantee learning convergence, capability of derivative, etc. On the other hand, a SVR is a novel method for tackling the problems of function approximation based on the statistical learning theory and has robust properties against noise. Hence, the CMAC-GBF systems combined the SVR approach to become the SVR-based CMAC-GBF systems are proposed. That is, the proposed structure has high accuracy in performance and noise against in the robust property. Finally, the experimental results demonstrate that the proposed systems outperform the CMAC-GBF systems.