A comparison study of forecasting financial fuzzy time series by artificial neural networks

Abstract

Forecasting financial time series is an important area for various investors and business practices. Fuzzy time series models which have been successfully applied to handle nonlinear problems have become popular in recent years for forecasting financial time series. Hence, this study aims to improve forecasting performance of financial fuzzy time series by using artificial neural networks. Istanbul stock exchange (ISE) national-100 index is used for forecasting. The empirical results show that fuzzy time series model which is applied in this study outperforms other forecasting models