Abstract

This paper is to predict economic development based on wavelet least squares support vector machine algorithm. The same as support vector machine, least squares support vector machine employs the principle of structure risk minimization, which has better generalization ability compared with SVM. In the study, wavelet kernel function can be used to create the kernel function of least squares support vector machine. In order to testify the effective of wavelet least squares support vector machine for economic development prediction, the comparison of the prediction error among the three prediction methods: wavelet least squares support vector machine, least squares support vector machine and support vector machine is performed. Obviously, the prediction error of wavelet least squares support vector machine for economic development is smaller than other prediction models. Therefore, we can conclude that the prediction ability of wavelet least squares support vector machine for economic development is better than least squares support vector machine and support vector machine.

Paper Details