基于ARFIMA模型的内蒙古农林牧渔业总产值研究

Based on ARFIMA Model Analysis of Inner Mongolia Agricultural Output

根据1947~2011年内蒙古农林牧渔业总产值的数据，运用时间序列分析对数据进行定量分析。以分数阶差分方法代替传统的整数阶差分方法，最大程度地保留了序列的原始信息；通过分数阶差分使数据平稳，经白噪声检验后，用 ARFIMA模型进行模型拟合；根据拟合模型对内蒙古农林牧渔业总产值进行预测，结果显示，模型的拟合结果较准确，误差在可控范围之内。

According to 1947-2011 Inner Mongolia agricultural output data.Time series analysis is used for quantitative analysis of the data.The traditional integer order difference method is replaced by fractional difference method,which retains the original information of the sequence in maximum extent. The fractional difference makes the data stationary.After the sequence is used for the white noise examina—tion,ARFIMA model is used to fit model,and test its significant.Finally according to the model Inner Mongolia the next two years agricultural output is predicted.The results show the fitting effects are more accurate,and the error is in the controllable range.