Study on improved grey integrated clustering method and its application

Abstract

The literatures about general grey clustering methods and many improved methods focus on the clustering results based on the principle of the maximal element of a grey clustering coefficient vector. Such methods neglect the effect of other elements except maximal one of a grey clustering coefficient vector to the cluster result. Because grey integrated clustering method comprehensively considers the effect of all elements of a grey clustering coefficient vector to the cluster result, it is an important development to grey clustering techniques. But the way that the value area of integrated clustering coefficients was divided into equal intervals needs improvement. This paper proposes an improved grey integrated clustering method, with which the value interval of a grey integrated clustering coefficient for a grey class was divided by regarding k±0.5 as boundary points. It is proved that when the difference of clustering coefficients of ordinary grey clustering methods is more than 1−1/(s−1), there are the same clustering results for the improved grey integrated clustering and the ordinary grey clustering. At last, we illustrate the improved grey integrated clustering method with the evaluation to rural economic development of Henan Province.