区域经济增长的混频预报与预测——以吉林省为例

The Nowcasting and Short-Term Forecasting of Regional Economic Growth——A Case of Jilin Province

将攫取的大量高频数据信息用于实际经济增长率的短期预测和实时预报,基于数据驱动的混频数据预测模型具有及时性、准确性和有效性的特征.该模型对吉林省实际经济增长率的预测与预报进行实证的结果表明:混频数据的自回归模型是非常有效的短期直接预测模型,混频数据模型的权重均值组合预测方法提高了预测结果的精确性和稳健性,吉林省经济正在进行有波动的筑底,"十三五"时期的经济运行将呈现稳中向好的态势.

Mixed data sampling model can grab lots of high-frequency data information for short-term forecasting and nowcasting,which is a data-driven model and has the characteristics of timeliness,accuracy and effectiveness.The model′s empirical results in real economic growth forecast of Jilin Province shows that mixed data sampling model with AR term is a very effective short-term direct forecasting model.The weight combination forecasting of mixed data sampling model significantly improves the prediction accuracy and robustness.The final results show that Jilin economic is reaching the bottomed with fluctuations,which means economic operation of Jilin province during the"13th Five-Year Plan"period will show a steady trend.