A new feature ranking criterion based on density function of subtractive clustering

Feature ranking is one of the basic methods in feature selection to select a subset of the original features. This paper uses a fuzzy clustering algorithm and proposes a new criterion for ranking the features. The importance of features is evaluated via the density function that is calculated in subtractive clustering. The proposed algorithm is tested over several well-known benchmark datasets. The performance of the proposed algorithm is also compared with some common algorithms. The results show that the proposed method is comparable to the other methods in term of obtained classification accuracy.