Exponential Length of Intervals for Fuzzy Time Series Forecasting


Bulut E., Duru O., Yoshida S.

IEEE Conference on Computational Intelligence for Financial Engineering and Economics (CIFEr), New-York, United States Of America, 29 - 30 March 2012, pp.107-112 identifier

  • Publication Type: Conference Paper / Full Text
  • City: New-York
  • Country: United States Of America
  • Page Numbers: pp.107-112

Abstract

The aim of this paper investigates the effective length of intervals for the fuzzy time series forecasting (FTSF) method. The length of intervals plays a significant role for the forecasting accuracy. The exponential length of intervals method is proposed and an empirical study is performed for forecasting of the time charter rates of Handymax dry bulk carrier ship. Rather than the existing literature, the proposed model is not only compared with the previous FTS models in which different length of intervals methods are applied, but also with the conventional time series methods such as the generalized autoregressive conditional heteroscedasticity GARCH model. The result of root mean squared error (RMSE) and mean absolute percentage error (MAPE) of proposed method is found superior than compared methods.