Neuro-fuzzy networks in time series modelling

The paper briefly presents and compares four neuro-fuzzy systems used for rule-based modelling of dynamic processes (chaotic Mackey-Glass time series). The following systems have been considered: nfMod, the system proposed in this paper; the well-known ANFIS and NFIDENT systems; and an alternative neuro-fuzzy system reported in literature. The main criterion of comparison of all systems is their performance (the accuracy of modelling) versus interpretability (the transparency and the ability to explain generated decisions; it also includes an analysis and pruning of obtained fuzzy-rule bases).