Fuzzy time series reflecting the fluctuation of historical data

The fuzzy time series is introduced by Song and Chissom to construct a pattern for time series with vague or linguistic value. Many methods using the interval and fuzzy logical relationship related with historical data have been suggested to enhance the forecasting accuracy. But they do not fully reflect the fluctuation of historical data. Therefore, we propose the interval rearranged method to reflect the fluctuation of historical data and to improve the forecasting accuracy of fuzzy time series. Using the well-known enrollment, the proposed method is discussed and the forecasting accuracy is evaluated. Empirical analysis shows that the proposed method in forecasting accuracy is superior to existing methods and it fully reflects the fluctuation of historical data.