Supervisory predictive control of weighted least square support vector machine based on Cauchy distribution

摘要

Least square support vector machine is a kind of thought to solve structural risk minimization method, weighted least squares support vector machine is introduced to solve the exist robustness, sparsity and large-scale computational problems, since the weighted method easily leads to shortcomings of over-fitting, according to the Cauchy distribution characteristics, weighted least squares support vector machines based on Cauchy distribution, and according to the identification function of least square support vector machine, which are used in supervisory predictive control algorithm. Simulation results show that weighted least square support vector machine based on Cauchy distribution learns fast, has good nonlinear modeling and generalization ability, and the supervisory predictive control algorithm of weighted least square support vector machine based on Cauchy distribution has better control performance.