A combination of fuzzy autoregressive integrated moving average model for forecast Malaysia's currency exchange rates

Abstract

Time series models such as Auto-Regressive Integrated Moving Average (ARIMA) models are the most important forecasting models used in financial market forecasting over the past three decades. However, the recent studies show that the combination ARIMA models with different kind of models can be an effective method of improving upon their predictive performance. Due to factors of uncertainty from the integral environment and rapid development of new technology, little data in a short time period to forecast future situations is encourage. This study is mainly concerned on the effectiveness of ARIMA and combination fuzzy ARIMA (F ARIMA) forecasting model, which combines the time series ARIMA model and the fuzzy regression model. Both of these models are applied in forecast the foreign exchange rates of Malaysian Ringgit to US dollars. F ARIMA model includes both interval models with interval parameters and the possible distribution of future value. Based on the results, it can be shown that the combination F ARIMA can makes a good forecast by using fewer observations than the ARIMA model.