Forecasting of Real GDP Growth Using Machine Learning Models: Gradient Boosting and Random Forest Approach
- PDF / 1,951,715 Bytes
- 19 Pages / 439.37 x 666.142 pts Page_size
- 123 Downloads / 306 Views
Forecasting of Real GDP Growth Using Machine Learning Models: Gradient Boosting and Random Forest Approach Jaehyun Yoon1 Accepted: 20 September 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This paper presents a method for creating machine learning models, specifically a gradient boosting model and a random forest model, to forecast real GDP growth. This study focuses on the real GDP growth of Japan and produces forecasts for the years from 2001 to 2018. The forecasts by the International Monetary Fund and Bank of Japan are used as benchmarks. To improve out-of-sample prediction, the cross-validation process, which is designed to choose the optimal hyperparameters, is used. The accuracy of the forecast is measured by mean absolute percentage error and root squared mean error. The results of this paper show that for the 2001–2018 period, the forecasts by the gradient boosting model and random forest model are more accurate than the benchmark forecasts. Between the gradient boosting and random forest models, the gradient boosting model turns out to be more accurate. This study encourages increasing the use of machine learning models in macroeconomic forecasting. Keywords Macroeconomic forecast · Random forest · Gradient boosting · Machine learning · Real GDP growth
1 Introduction The ability to forecast macroeconomic variables is highly desirable for the design and implementation of timely policy measures. Among the macroeconomic variables, real GDP growth is one of the most important data. However, forecasting real GDP growth involves complicated calculations, and official data are often available only after at least a one-quarter delay. Due to this delay, policymakers often design and implement policies without knowing the necessary information.
* Jaehyun Yoon [email protected] 1
Faculty of Political Science and Economics, Waseda University, 1‑6‑1, Nishi‑Waseda, Shinjuku‑ku, Tokyo 169‑8050, Japan
13
Vol.:(0123456789)
J. Yoon
From this point of view, if available, the accurate forecasting of real GDP growth in advance would be highly valuable. Forecasting macroeconomic data, such as real GDP growth, is not a simple process. To forecast data, considering the causal relationship between the dependent variable and independent variable, traditional economic forecasting models require predetermined relevant variables to make predictions and often take top-down and theory-driven approaches (Mullainathan and Spiess 2017). This process also requires economic intuition and judgment by forecasters regarding the data and methods used. If there is any flaw in the assumptions made by the forecasters, the models could produce inaccurate predictions. In contrast to many traditional economic forecasting models, machine learning models mostly deal with pure prediction (Varian 2014). Machine learning models are more flexible than traditional economic forecasting models and can produce predictions without predetermined assumptions or judgments. In fact, in conjunction with technol
Data Loading...
