A learning algorithm of least squares support vector machine based on factor analysis and Renyi-entropy

By comparing and analysing the algorithm of least squares support vector machine (LS-SVM) based on Renyi-entropy, traditional least squares support vector machine and standard support vector machine(SVM), this paper concludes whether the number of training samples or training time, LS-SVM model based on Renyi-entropy are significantly better than the model of traditional LS-SVM and standard support vector machine.In addition, through the factor analysis treatment, multicollinearity between the original input variables are effectively eliminated. Although the number of input variables are decreased, still achieve a higher prediction accuracy rate, which provides that these public factor contains the majority of financial information and the method of factor analysis in this paper is effective.