指数平滑预测法及其在经济预测中的应用

摘要

以1978—2008年的某省职工年平均工资作为样本数据、2009—2010年的数据作为模型检验数据，建立基于时间序列分析的指数平滑预测模型.检验结果表明，指数平滑预测模型对2009年和2010年的预测值与检验样本的实际值之间的相对误差很小(仅为0.015032和0.02207)，预测结果理想.随后，以此模型预测时间序列2012—2015年职工平均工资数据.%From 1978 to2008annual average wages of staff and workers of the province as the sample data,2009 to 2010 data as a model test data, based on the time series analysis of exponential smoothing prediction model.The test results showed that, exponential smoothing prediction model on2009 and2010 forecast value and the actual value of the test sample between the relative error is very small (only 0.015032 and 0.02207), prediction of results of ideal.Subsequently, this model to forecast the time series data of 2012—2015 average wage of staff and workers.