Enhanced Type 2 Fuzzy System Models With Improved Fuzzy Functions

Abstract

A new Fuzzy System Modeling (FSM) approach based on Improved Fuzzy Functions using Discrete Interval Type 2 Fuzzy Sets is presented. The new method is proposed as an alternate learning and reasoning schema to Type 1 and Type 2 FSM with Fuzzy Rule Base (FRB) approaches and enhances Type 2 FSM by reducing complexity and increasing prediction performance. Structure identification of the new approach is based on a supervised Improved Fuzzy Clustering (IFC) method with a dual optimization algorithm, which yields improved membership values. The merit of the proposed Type 2 FSM is that uncertain information on natural grouping of data samples, i.e., membership values, is utilized as additional predictors while structuring fuzzy functions. The uncertainty in selection of the learning parameters are captured by identifying two separate features: executing IFC method with varying levels of fuzziness values, m, and collection of different fuzzy function structures. It is shown with an empirical study that the new Type 2 FSM approach is superior in comparison to earlier Type 1 and Type 2 FSMs in terms of robustness and error reduction.