Type-2 Fuzzy Sets for Pattern Classification: A Review
Jia Zeng, Zhiqiang Liu · 2007
This paper reviews the advances of type-2 fuzzy sets for pattern classification. The recent success of type-2 fuzzy sets has been largely attributed to their three-dimensional membership functions to handle more uncertainties in real-world problems. In pattern classification, both feature and hypothesis spaces have uncertainties, which motivate us of integrating type-2 fuzzy sets with traditional classifiers to achieve a better performance in terms of robustness, generalization ability, or classification rates. We describe recent type-2 fuzzy classifiers, from which we summarize a systematic approach to solve pattern classification problems. Finally, we discuss the trade-off between complexity and performance when using type-2 fuzzy classifiers, and explain the current difficulty of applying type-2 fuzzy sets to pattern classification