FORECASTING OF TURKEY'S UNEMPLOYMENT RATE FOR FUTURE PERIODS WITH ARTIFICIAL NEURAL NETWORKS

Abstract

In world economies, in order to achieve high national income level, employment has an important effect. Therefore, it is necessary for unemployment to be highly low. Labor force structure of a country specifies the state of that country, and that unemployment rates are at high levels is an indicator of that there is a deviation in the development and growth rate of country economy. In this context, forecasting unemployment rates in the next periods of Turkey, in order to develop solution suggestions for unemployment problem which is one of the most important problems of today, and make contribution to improving country economy, it was decided to conduct such a study. In this forecasting study conducted, due to the fact that the risk to obtain wrong results with traditional methods is high in the solutions of chaotic contented problems, it was decided to be used ANN (Artificial Neural Network) method, which presents healthy solutions of the chaotic problems including partly erroneous or over deviating data and is one of the contemporary methods, commonly used in the literature. In the study, utilizing the monthly basic economic indicators of Turkey belonging to the period of 2005-2018, forecast of unemployment rate for the next periods was made by ANN method, and the data belonging to totally six basic economic indicators were used in the forecast. As a conclusion of the study, it was identified that the forecast made by the model developed produced the reliable results that are quite close to the reality.