A new evidence accumulation method with hierarchical clustering

Evidence accumulation clustering is widely used with the developing of data mining. With the coming of the era of big data, it is inevitable that the clustering will be used in big data. This paper proposed a new evidence accumulation clustering with hierarchical clustering method. It clusters a feature one by one rather than clustering all features at the same time. The algorithm uses voting mechanism to compose a co-association matrix and uses hierarchical clustering algorithm with single link to get the final partition. The proposed method is tested on KDD99 dataset and 8 other datasets. And it could achieve better results than evidence accumulation clustering method and save computation time.