Analysis of the Phillips Curve for Turkiye: A Comparison of the Johansen Cointegration and Artificial Neural Network Models
Makaleler > Tam Makale · 2024
Yazarlar: Dilek Surekci Yamacli, Çağatay TUNÇSİPER
INTERNATIONAL JOURNAL OF CONTEMPORARY ECONOMICS AND ADMINISTRATIVE SCIENCES, ESCI İndekslerine Giren Dergi, Uluslararası, Wos, 2024, cilt 14, ss 87-106
Özet
This study is examined the validity of the Phillips curve hypothesis using the Johansen cointegration and artificial intelligence methods for the period of 2010-2022 in T & uuml;rkiye and it is also compared the forecasting performance of these methods using coefficient of determination values and error values. Results of the study an 1% increase in unemployment in the long run leads to a 0.854% decrease in consumer prices. This result supports the Phillips Curve hypothesis for the long run. On the other hand, the error correction model shows that Phillips curve hypothesis is not valid in the short run. Besides, R2 and other error metrics namely mean absolute error, root mean square error and the mean absolute percentage error values verify the better forecasting performance of the artificial neural network model than Johansen cointegration. In this context, the findings that can be obtained as a result of artificial intelligence modeling in the management of inflation and unemployment are considered to be important.