Presentation of a new hybrid approach for forecasting economic growth using artificial intelligence approaches (original) (raw)

2019, Neural Computing and Applications

This study presents a new hybrid algorithm for forecasting economic growth using indicators of knowledge-based economy (KBE). The algorithm consists of three steps, namely preprocessing, processing, and postprocessing. Preprocessing consists of principal component analysis and reproduction algorithm, which are used to decrease the number of variables and increase the volume of data. Economic growth is predicted during processing using multilayer perceptron (MLP), adaptive neuro-fuzzy inferences system, and gene expression programming (GEP). The variables are added separately to the process. The best model is selected during the postprocessing step to forecast economic growth. GEP model is used to forecast unique indicators. The last step involves substitution of forecasted indicators in the best model. In this study, the KBE indicators of Iran from 1993 to 2013 are predicted in the processing step. The MLP model is used, which includes four indicators, namely technological foundation, structure of trained manpower, export and trademark, and employee. Indicators are also forecasted using the GEP model between 2013 and 2020. The results are used to estimate economic growth between forecasting periods. A self-organizing map is used to recognize relationships between variables. The results show the efficiency of the algorithm in multivariate forecasting.

Sign up for access to the world's latest research.

checkGet notified about relevant papers

checkSave papers to use in your research

checkJoin the discussion with peers

checkTrack your impact