« Back to overview
Artificial neural networks have become increasingly popular for statistical model fitting over the last years, mainly due to increasing computational power. In this paper, an introduction to the use of artificial neural network (ANN) regression models is given. The problem of predicting the GDP growth rate of 15 industrialized economies in the time period 1996-2016 serves as an example. It is shown that the ANN model is able to yield much more accurate predictions of GDP growth rates than a corresponding linear model. In particular, ANN models capture time trends very exibly. This is relevant for forecasting, as demonstrated by out-of-sample predictions for 2017.
The main goal of the paper is to demonstrate the capabilities of artificial neural network (ANN) regression models. In particular, the paper shall encourage researchers to use ANN models for regression analysis. It shall help to overcome two existing barriers to the use of ANN models: the lack of theoretical understanding on the one hand and the inability to actually implement the model on the other hand. Therefore, the theoretical framework is discussed along with the precise algorithm used to train the ANN. The academic usefulness is demonstrated by the ability of the ANN regression model to produce more accurate predictions of GDP growth rates of 15 developed countries between 1996 and 2016 than a corresponding linear model. Regarding forecasts, the exible time trend is shown to be a major advantage of the ANN model.
Keywords: neural network, forecasting, panel data
JEL codes: C45, C53, C61, O40
The INTERREG project DUAL Ports is concerned with the development of innovative ways to reduce the carbon emissions in and around small and medium-sized entrepreneurial ports in the North Sea Region.MORE
Political decisions have influenced prices on the international commodity markets in this year.MORE
The BDO International Business Compass (IBC) aims to present the overall business development status of countries and regions in the form of a single index value. MORE
Artificial neural network regression models: Predicting GDP growth