基于GIOWA算子的我国货币需求水平组合预测研究

On the Combination Prediction of China’s Currency Demand Level Based on GIOWA Operator

基于1995—2020年的我国广义货币发行量数据,依据多元回归模型、ARIMA(1,1,2)模型和HoltWinters无季节模型,构造以误差平方和最小为目标的GIOWA算子组合预测模型,对我国2021—2023年的货币需求水平进行预测,并建立误差评价体系对该模型进行有效性评价.结果表明,2021—2023年我国货币需求量整体呈现出上升的趋势;诱导因子为1时的组合预测模型的表现效果最好;货币需求的增长幅度会是一个先上升后下降的过程.

In order to predict the level of money demand in my country in the next three years, based on my country’s broad currency issuance data from 1995 to 2020, We will establish a GIOWA operator combination prediction model with the target of minimizing the sum of square of the error, based on multiple regression models, ARIMA(1, 1, 2) models and Holt-Winters no season model. At the same time, establishing an error evaluation system to evaluate the model. The study finds that the GIOWA combined forecasting model can better forecast my country’s money demand. In the next three years, my country’s money demand will show an upward trend as a whole. The combined forecasting model with an inducing factor of 1 performs best;the growth rate will be a process of rising first and then falling.