An evolutive Interval Type-2 TSK Fuzzy Logic System for volatile time series identification

This paper presents a study case of the US Dollar -Colombian Peso Exchange Rate by using a First-order Interval Type-2 TSK Fuzzy Logic System. This study case is specially interesting because it presents a volatile behavior which is a complex problem for classical analysis. Their results are checked by statistical tests as Bayesian, Akaike, Hannan-Quin criteria, Goldfeld-Quant, Ljung-Box, ARCH, Runs and Turning Points which provide appropriate criterions to test the solution. Some methodological aspects about a hybrid implementation between Evolutive Optimization and First order Interval Type-2 TSK FLS are presented. Additionally, the selected type-reduction algorithm is the IASCO algorithm proposedby Melgarejo in since it presents better properties than other algorithms.