Forecasting economic growth at provincial level in Vietnam: A systematic dynamics modeling approach

Forecasting annual economic growth is an important task to help local governments set goals and policies for socio-economic development. This study proposes and develops the forecasting method for economic growth at the provincial level in Vietnam named Dynamic Systems Modeling. It is an effective method to analyze the dynamic interactions between economic growth and related factors, which is useful for forecasting economic growth. This method is a combination of traditional statistics and machine learning that brings many advantages in forecasting. Specifically, (1) This method is used to forecast panel data, which helps to control both temporal and spatial problems of the forecast object; (2) This forecasting method bases on the comparison and selection of many different models; (3) Forecast results are verified, ensuring reliable forecast results. Through the data set collected from the Vietnam General Statistics Office from 2016 to 2020, we apply the Dynamic Systems Modeling method to forecast two important economic growth indicators at the provincial level, which are GRDP and GRDP per capita. The analysis and forecasting evaluation results show that the Dynamic Systems Modeling is an effective tool for forecasting economic growth at the provincial level in Vietnam.