A Regime Switching Model for Fuzzy Time Series Forecasting
This paper proposes a regime switching model to improve fuzzy time series forecasting. Conventional fuzzy time series models create only single set of fuzzy relationships for forecasting. When we forecast economic and financial time series, which may face drastic changes, the forecasts from these models may not be satisfactory. Hence, the application of these models may be limited. The regime switching model can cluster the time series into different regimes. Each regime may become a time series. We then can conduct fuzzy forecasting for each regime.Daily stock index will he used for empirical analyses.
clusters fuzzy relatinships stock index
Kun-Huang Huarng Tiffany Hui-Kuang Yu
Dept. International Trade Feng Chia University Taiwan Dept. Public Finance Feng Chia University Taiwan
国际会议
The 10th International Conference on Intelligent Technologies(第十届智慧科技国际会议 InTech09)
桂林
英文
242-246
2009-12-12(万方平台首次上网日期,不代表论文的发表时间)