Identification in Takagi-Sugeno Fuzzy Models

In this paper, we evaluate the parameters and study the applicability of the Takagi-Sugeno model to the forecasting of stock market indicators by the example of the main indices of the Moscow stock exchange: the MICEX index, the RTSI index, and the oil and gas industry index. At the beginning of the work, a literature review was prepared have a closer look at application of the TS-model for predicting stock indices in order to improve the quality of the forecast. In light of this articles, we evolve several econometric approaches to the study of the behavior of stock indices, which formed the basis of basic designs in the modeling of fuzzy rules in this paper, were singled out. A comparative analysis of the influence of the t-norm on the quality of the assessment and the coefficients before the fuzzy rules was also carried out. Based on the results of the work, the Takagi-Sugeno model showed a decrease in the forecast error, which can reach 3-5 times in comparison with the regression approach, depending on the set of independent variables. However, the results of the TS-model with fuzzy rules based on the time series turned out to be slightly worse than for a similar AR-process. In general, the model showed better results than traditional methods of forecasting.