A Rough-fuzzy C-means using information entropy for discretized violent crimes data

This paper presents the factor clustering analysis for violent crimes. The efficiency of Rough-fuzzy C-means algorithm is affected by the numbers of clusters, and not all centroids are beneficial. The analyzing of violent crime data does not need human intervention for impartiality. The information entropy is a helpful tool for resolving those issues. In this paper, a novel discrete Rough-fuzzy C-means based on information entropy algorithm (DRFCMI) is proposed, which can obtain typical conclusions objectively. Experimental results illustrate that our proposed method is efficient.