Abstract:

A modeling method of the Mamdani type probabilistic fuzzy system is presented to deal with randomness residing in fuzzy systems. Probabilistic fuzzy rules are directly generated from numerical data in this model without the aid of domain experts. A probabilistic fuzzy reasoning scheme is proposed accordingly. The probabilistic fuzzy system can provide better representation of randomness residing in data, by combining the interpretability of fuzzy systems with the statistical properties of probabilistic systems. The application to the analysis of time series is taken as an example to demonstrate the effectiveness of the proposed method through a comparison with the table look-up scheme.