Abstract
Support Vector Machines (SVM) are new machine learning methods, and are studied about many applications. However, a regression method of SVM shows little availability. In this paper, we propose a learning control scheme using Support Vector Regression (SVR). This control system consists a pre-learned SVR, and is applied to a control problem of inverted pendulum. SVR system learned input-output data of state feedback control given an initial condition. The computer simulation results show the availability of the proposed control system.
