准确预测国际原油价格,对于维护经济稳定和规避风险具有重要意义.由于国际油价的波动是由多种因素引起的,本文采用误差修正模型确定原油价格与因素的关系.将结果作为SVM回归预测模型的输入模式,建立基于多因素SVM的油价预测模型.通过实证研究,发现基于多因素SVM的油价预测模型相对于误差修正模型和基于国际油价本身的自回归SVM预测模型具有更好的拟合和预测效果.
It is a great significance to forecast the international petroleum price for mainte- nance of economic stable and dodge of risk. The fluctuation of international petroleum price is caused by multi-faceted factors, the relation between petroleum price and factors is defined by the error amendment model. The result is as the input of the SVM regression model and the forecast model for petroleum price base on the multi-faceted factors SVM is builded. Ac- cording to the study, the forecast model for petroleum price base on the multi-faceted factors SVM have the better effect of fitting and forecast than the model of error revised and the model of autoregression SVM forecast model based on international petroleum price.