Sparse online LS-SVM based on modified particle swarm optimization algorithm and application

In this paper, A simple and effective mechanism is proposed to realize the parsimoniousness of the online least squares support vector machine. Hence, the response time is curtailed. Besides, a modified Particle Swarm Optimization (PSO) algorithm is proposed to ascertain the optimal model parameters. Simulation results show that it outperforms GA and common PSO algorithms and LS-SVM model based on the modified PSO algorithm has the best regression accuracy and generalization ability. The sparse online LS-SVM algorithm is applied to build a turbine heat rate forecasting model, which possesses dynamic prediction functions.