The impact of FOU size and number of MFs on the prediction performance of Interval Type-2 Fuzzy Logic Systems

Abstract

The inclusion of footprint of uncertainty (FOU) in Interval Type-2 Fuzzy Logic Systems (IT2FLSs) made them suitable for modelling uncertainty. This paper investigates the impact of FOU size and number of membership functions (MFs) on the model's prediction performance. An IT2FLS trained using a fast learning method is designed here. The uncertainty in data is captured by designing the IT2FLS with different sizes of FOU. The concept of extreme learning machine (ELM) is then used for optimal tuning of IT2FLS consequent parameters. The designed model is applied to the chaotic time series prediction. During simulation it is observed that the increase in FOU size with the increase in number of MFs give better prediction results.