?
Применение моделей, основанных на нечеткой логике, к финансовым временным рядам
The generalized autoregressive conditional heteroscedasticity model is widely applied to financial time series. There are further generalizations of this model. One of such generalizations is a combination of Takagi–Sugeno type fuzzy systems and autoregressive conditional heteroscedasticity models. The Takagi–Sugeno fuzzy systems advantage is that there is a standalone generalized autoregressive conditional heteroscedasticity model constructed for each fuzzy cluster (for example, a cluster “low volatility, moderate return”). Due to calculation of fuzzy rule activation level on a specific part of time series the soft switching between these models is provided. The topic discussed in this report is the position of Takagi–Sugeno fuzzy systems in the fuzzy logic field. Some results of calculations for Russian time series are also presented.