集成灰色支持向量机预测模型研究与应用

Abstract

对灰色预测gM(1,1)模型进行了分析,提出了集成灰色支持向量机的预测模型。分别对影响灰色预测gM(1,1)模型精度的背景值的计算、初值的选取以及数据序列的光滑度进行改进,提出了背景gM模型、初值gM模型、光滑度gM模型,并结合支持向量机的特点,将一维原始数据序列通过三个灰色模型得到的三组值作为支持向量机的输入,原始序列作为支持向量机的输出,训练得到最佳支持向量回归机模型。仿真结果表明了该模型的有效性。

Based on grey prediction GM (1,1) model,an integrated grey Support Vector Machine (SVM) model was presented.Through improving the GM (1,1) prediction accuracy based on background value calculation,initial value selection and smooth degree of data sequence,three grey prediction models that are background GM model,initial value GM model,smooth degree GM model,were put forward.Then,combining the advantages of SVM,the prediction results of three grey prediction models were used as the SVM input factor,and the original data sequence was used as the output factor of the SVM.The support vector regression machine was trained to get the optimal structure.The results of experiment show that the model is valid.