ARIMA与SVM组合模型的石油价格预测  

Oil Price Forecasting Based on ARIMA and SVM Hybrid Model

作　　者：吴虹[1] 尹华[1] 

机构地区：[1]赣南师范学院数学与计算机科学学院,江西赣州341000

出　　处：《计算机仿真》2010年第5期264-266,326,共4页Computer Simulation

基　　金：国家自然科学基金(30570352)资助

摘　　要：针对复杂时间序列预测困难的问题,在综合分析其线性和非线性复合特征的基础上,提出了一种基于ARIMA和SVM相结合的时间序列预测模型。首先采用ARIMA模型对时间序列进行线性建模,然后采用SVM对时间序列的非线性部分进行建模,最后得到两种模型的综合预测结果。将组合模型应用于石油价格预测中,仿真结果表明组合模型相对于单模型的预测具有更高的精度,发挥了2种模型各自的优势,在复杂时间序列预测中具有广泛的应用前景。In order to solve the problem of complex time series forecasting including the linear and nonlinear features, a new hybrid forecasting model based on ARIMA and SVM is proposed in this paper. ARIMA model was used to predict the linear component of complex time series and SVM model was applied to the nonlinear residual component, and the hybrid forecasting results were obtained. The prediction performances of the methods are tested on simulation experiment for oil price. The results show that the hybrid model, which takes advantage of the unique strength of the two models in linear and nonlinear modeling, has better accuracy than the single model. The hybrid model is an effective method for complex time series.

关 键 词：支持向量机 差分自回归移动平均 组合预测 石油价格 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]