Determining the Necessary Criteria for the EU Membership by Using the Machine Learning Method: 2005-2010 Period Analysis

Abstract

In recent years, providing the required criteria put forward for the full membership by the union is one of the prior targets of Turkey in the European Union full membership process. The objective of this study is to be able to evaluate situations of the membership process among European countries, candidate countries and Turkey. In the data set used in the period of 2005-2010, it is seen that the global economic crisis becoming efficient towards the end of 2007 has affected both EU countries and candidate countries economically. This period which included the effect of global crisis was taken as a data set to the study in order to add the effect of global crisis into the evaluation. The study was tried to be analyzed by using artificial neural networks and decision trees algorithms by means of a variety of economic signs. For that purpose, the data between years of 2005-2010 were used in this study. Primarily, the relative importance of independent variables was found by means of multilayered artificial neural networks and then dimension reduction operation was done. The criteria thought as affecting the membership was tried to be found out by using the decision trees method. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility 7th International Strategic Management Conference