ARIMA模型在贵州省农产品价格预测中的应用——以辣椒为例

摘要

由于贵州省辣椒月价格呈现季节性,因此以2007年1月至2010年12月贵州省辣椒月价格为例,在运用季节分解方法剔除其季节因素形成新序列的基础上,构建了非平稳时间序列ARIMA(p,d,q)模型,并预测了辣椒未来的月价.结果表明,拟合指标优良的ARIMA(1,1,1)模型能很好地预测辣椒月价格趋势,并将辣椒价格的预测值与实际值的相对误差基本控制在9％以内,实证分析结果证明了ARIMA模型的季节分解方法在贵州省农产品价格预测中的预测性和可行性.%Because the seasonal nature of monthly price of hot pepper in Guizhou Province, therefore, taking the case of monthly price of hotpepper in Guizhou Province from January in 2007 to December in 2010, seasonal decomposition method was used to reject the new sequenceformed by seasonal factors, nonstationary time series ARIMA(p,d,q) Model was constructed to predict the monthly price of hot pepper. Theresults indicated that the ARIMA (1,1,1) Model could predict the peppers price trend, and the difference between the forecast price and theactual price was more or less than 9%. The empirical results indicated the predictability and feasibility of seasonal decomposition method ofARIMA Model in the prediction of the price of agricultural products in Guizhou Province.