Comparison of Four Interval ARIMA-base Time Series Methods for Exchange Rate Forecasting

In today’s world, using quantitative methods are very important for financial markets forecast, improvement of decisions and investments. In recent years, various time series forecasting methods have been proposed for financial markets forecasting. In each case, the accuracy of time series methods fundamental to make decision and hence the research for improving the effectiveness of forecasting models have been curried on. In the literature, Many different time series methods have been frequency compared together in order to choose the most efficient once. In this paper, the performances of four different interval ARIMA-base time series methods are evaluated in financial markets forecasting. These methods are including Auto-Regressive Integrated Moving Average (ARIMA), Fuzzy Auto-Regressive Integrated Moving Average (FARIMA), Fuzzy Artificial Neural Network (FANN) and Hybrid Fuzzy Auto-Regressive Integrated Moving Average (FARIMAH). Empirical results of exchange rate forecasting indicate that the fuzzy artificial neural network model is more satisfactory than other models. Index Terms: Artificial Neural Networks (ANNs), Time series forecasting, Auto-Regressive Integrated Moving Average (ARIMA), Combined forecast, Exchange Rate.