Weighted support vector echo state machine for multivariate dynamic system modeling

Support vector echo state machine (SVESM) is a promising dynamic system modeling tool, which performed linear support vector regression (SVR) in the high dimension “reservoir” state space. A variant of SVESM, weighted support vector echo state machine (WSVESM) is proposed in this paper to deal with the multivariate dynamic system modeling problem. The historical observed data of the dynamic system are treated as multivariate time series, and the proposed WSVESM model is used to predict the time series. Different weights are allocated to the training data, and a multi-parameter solution path algorithm is introduced to determine the solution of WSVESM. Simulation results based on artificial and real-world examples show the effectiveness of the proposed method.