General Information

The objective of this paper is to forecast quarterly GDP in Macao using different neural network models. It is a challenge task due to the scarcity of determinant economic indicators and the scarcity of economic data. We compared the forecast errors of three different neural network models including Back Propagation (BP), Elman and Radial Basis Function (RBF). Elman has never been used in the GDP forecasting in literature, however in our results, Elman has the least forecasting error due to its recurrent network topology which can remember the history economic data.