摘要
【目的】针对林业产值预测,提出一种基于季节性时间序列的方法,为林业部门制定合理规划提供重要依据。【方法】对林业产值序列采用CH检验方法检验其平稳性,考虑序列的周期性和趋势性特点,建立基于季节性时间序列的林业产值预测模型,并与传统的GM(1,1)模型进行对比分析。【结果】与传统的GM(1,1)模型相比,季节性时间序列模型不仅可以预测以年度数据为单位的经济序列,而且可以预测以季度数据和月度数据为单位甚至是每日数据为单位的经济序列,且具有更高的预测精度。【结论】季节性时间序列模型可有效地预测林业产值的周期性和趋势性,且可大幅度提高预测精度,可用于林业部门对林业产值的预测和规划。
【Objective】Consider that periodic performance can improve the fitted effect of the model of forestry output value and predict the output value of forestry in the future,which can provide a reference for subsequent forestry departments to implement rational planning. 【Method】Selecting data for the period 2004—2016,total output value of forestry,in view of the periodicity and trend of forestry output value and combining the traditional ARIMA model,we have designed a time series model of seasonal forestry output value. The model uses the CH test method to make up for the shortage of the traditional GM（ 1,1） model on the seasonal data test. 【Result】It is found that the seasonal time series model can not only predict the economic sequence based on annual data,but also can predict the economic sequence of the quarterly data and monthly data units or even daily data with higher prediction accuracy compared with the traditional GM（ 1,1） model.【Conclusion】It can be used by the forestry department to plan forestry output value because a seasonal time series model can effectively predict the periodicity and trend of forestry output value,and greatly improve the predictive accuracy.
作者
沈杰
杨忠月
乔吉良
SHEN Jie;YANG Zhongyue;QIAO Jiliang(College of Economy and Management,Nanjing Forestry University,Nanjing 210037,China)
出处
《南京林业大学学报（自然科学版）》
CAS
CSCD
北大核心
2018年第5期185-190,共6页
Journal of Nanjing Forestry University：Natural Sciences Edition
