Economic forecasting based on chaotic optimized support vector machines

The economic system, especially the macroeconomic system, is a complex system with nonlinear, time-varying and coupling characteristics. Aiming at the macroeconomic modeling and forecasting problem, a support vector machine method is proposed in this paper. The modeling method of least square support vector machine is mathematically analyzed first, and then an improved multi-scale chaotic optimization algorithm combined with the genetic algorithm is proposed to optimize the model parameters. Using historical economic data, the model is trained and used for forecasting. Forecasting results show that the prediction accuracy has been improved, the average error rate decreases from 15% achieved by the BP neural network to less than 4% by the proposed algorithm.