Membership Functions Effect for Creating Fuzzy Classifiers Ensemble

Ensemble of classifiers is a new method to achieve more accurate classifiers. Diversity among the members of a team of classifiers is deemed to be a key issue in classifiercombination. However, measuring diversity is not straightforward because there is no generally accepted formal definition. There are several methods for measuring diversity.In this paper we try to measure diversity in fuzzy classifiers ensemble versus different fuzzy membership functions. Ournumerical result shows that using triangular membership function in fuzzy classifiers is better than others

کلیدواژه ها:

Diversity ، Ensemble of classifiers ، fuzzy classifiers ، membership functions