Forecasting Neural Network-Based Fuzzy Time Series with Different Neural Network Models

Fuzzy approach and artificial neural networks become effective tool for researchers by forecasting fuzzy time series. The relation of these has advantage to improve forecasting performance especially in handling nonlinear systems. Hence, in this study we aimed to handle a nonlinear problem to apply neural network-based fuzzy time series model. Differing from previous studies, we used various degrees of membership in establishing fuzzy relationships and we performed different neural network models to improve forecasting performance. To demonstrate comparison between these models we used a data set of exchange rate of Turkish Liras (TL) to Euro for the years 2005-2009. Empirical results show that the multilayer perceptron is the best to forecast fuzzy time series in most commonly used artificial neural network models. Key-Words: Fuzzy approach, neural networks, forecasting, time series, exchange rate, degrees of membership.