Fuzzy feature selection based on min–max learning rule and extension matrix

In many systems, such as fuzzy neural network, we often adopt the language labels (such as large, medium, small, etc.) to split the original feature into several fuzzy features. In order to reduce the computation complexity of the system after the fuzzification of features, the optimal fuzzy feature subset should be selected. In this paper, we propose a new heuristic algorithm, where the criterion is based on min–max learning rule and fuzzy extension matrix is designed as the search strategy. The algorithm is proved in theory and has shown its high performance over several real-world benchmark data sets.