Applying Lasso Linear Regression Model in Forecasting Ho Chi Minh City’s Public Investment

Forecasting public investment is always an issue that has attracted much attention from researchers. More specifically, forecasting public investment helps budget planning process take a more proactive approach. Like many other economic variables such as oil prices, stock prices, interest rates, etc., public investment can also be forecasted by different quantitative methods. In this paper, we apply the Ordinary Least Square (OLS), Ridge, and Lasso regression models to forecasting Ho Chi Minh City’s future public investment. The most effective forecasting method is chosen based on two performance metrics – root mean square error (RMSE) and mean absolute percentage error (MAPE). The empirical results show that the Lasso algorithm has superiority over the two other methods, OLS and Ridge, in forecasting Ho Chi Minh City’s Public Investment.