Research On Fuzzy Multiple Output Least Squares Support Vector Machine Classification And Regression

Machine learning has become a hot research topic.The support vector machine(SVM)based on statistical learning theory in machine learning method can solve practical problems such as small sample,nonlinear and local minima.However,there are many problems in the training process of SVM,such as noise and long learning time.In this paper,fuzzy least squares support vector machine(FLS-SVM)and multiple output intuitionistic fuzzy least squares support vector regression(IFLS-SVR)are proposed to solve these problems,and the multiple output intuitionistic fuzzy least squares support vector regression is applied to wind weather prediction.In order to improve the training efficiency and reduce the training time of SVM,least squares method is introduced.In order to solve the problem of noise and field in the support vector regression model,a fuzzy least squares support based on fuzzy theory is proposed.An improved membership calculation method is proposed,which takes into account not only the distance between the sample point and the class center,but also the distance between the sample points and the classification plane.And the bat algorithm is used to optimize the parameters of the kermel function to improve the limitations and shortcomings of the algorithm.Experimental results on artificial datasets show that the FLS-SVM algorithm has good classification performance and is robust to random noise.The multiple output intuitionistic fuzzy least squares support vector regression introduces intuitionistic fuzzy on the basis of multiple output SVM to solve the problem of uncertain multiple output complex system.Compared with the traditional fuzzy support vector regression,it is closer to the actual model except the fuzzy membership degree and the non-fuzzy membership degree.The multiple output IFLS-SVR transforms the actual data into fuzzy data by intuitionistic fuzzy algorithm,and transforms the two programming optimization problem into solving a series of linear equations.Compared with the existing fuzzy support vector regression,the multiple output IFLS-SVR uses the intuitionistic fuzzy method to calculate the membership function.The least squares method is used to improve the training efficiency,reduce the training time and get a more accurate solution.Through simulation model,the multiple output IFLS-SVR achieves good results compared with other methods.Multiple output intuitionistic fuzzy least squares support vector regression is applied to complex wind forecasting.By calculating the correlation degree of wind meteorology,selecting the relative factor of relative degree,and the multiple time series of wind speed and wind direction,the related data are extracted from the multiple time series.The prediction model is established and the experiment is carried out,and the better prediction results are obtained.