Neural Networks-Based Forecasting Regarding the Convergence Process of CEE Countries to the Eurozone

Magdalena RĂDULESCU, Logica BANICA


In the crisis frame, many forecasts failed to provide well determined ratios. What we tried to explain in this paper is how some selected Central and Eastern European countries will perform in the near future: Romania, Bulgaria, Hungary, Poland and Czech Republic, using neural networks- based forecasting model which we created for the nominal and real convergence ratios. As a methodology, we propose the forecasting based on artificial neural network (ANN), using the well-known software tool GMDH Shell. For each output variable, we obtain a forecast model, according to previous values and other input related variables, and we applied the model to all countries. Our forecasts are much closer to the partial results of 2013 in the analyzed countries than the European Commission’s or other international organizations’ forecasts. The results of the forecast are important both for governments to design their financial strategies and for the investors in these selected countries. According to our results, the Czech Republic seems to be closer to achieve its nominal convergence in the next two years, but it faces great difficulties in the real convergence area, because it did not overpass the recession.


nominal convergence; real convergence; euro-zone; Central and Eastern European Countries; forecasting model; neural networks

Full Text:

Creative Commons License
Transylvanian Review of Administrative Sciences by TRAS is licensed under a Creative Commons Attribution 4.0 International License.
Based on a work at

Online ISSN: 2247-8310 | Print ISSN: 1842-2845 |  © AMP

The opinions expressed in the texts published are the author’s own and do not necessarily express the views of TRAS editors. The authors assume all responsibility for the ideas expressed in the materials published.