Fuzzy modeling in stock-market analysis

The article examines the application of Takagi-Sugeno fuzzy models (T. Takagi and M. Sugeno, 1985) to the problem of stock market analysis. Different model structures are evaluated in a case study on the modeling of the Dutch AEX-price index. A scenario model is used for examining "what-if" scenarios and a prediction model searches for predictive components in relevant (macro) economic variables. It is found that TS models can be applied successfully in these areas, due to their capability of approximating general nonlinear systems and to their transparency. Further research is recommended.