No abstract

In some cases, clustering objects into several compatible clusters is more rational than traditional clustering method5 do. In this paper, we propose a new compatible clustering algorithm based on CompClustering[8j, it adopts point neighborhood technique to replace the iterative mechanism of the latter. Experiments show that the proposed algorithm can get some consistent clustering results, and theory analysis also demonstrates that the proposed algorithm has lower computation consumption than CompClustering does.

No abstract

Dissimilarity measure plays a very important role in traditional data clustering. In this paper, we extend the dissimilarity measure as compatible measure and present a new algorithm (CNclustering) based on this measure. The algorithm is a rigorous partition method, it first gets some compatible clusters with a Compclustering method as the initial nucleoids, then absorbs other objects by the absorbing step to form the final clusters. We use S20 and S200 data sets to demonstrate the clustering performance of the algorithm and get some consistent results。 Index Terms-clustering algorithm, dissimilarity, nucleoid, compatible relation, absorbing.I.

Special clustering algorithms are attractive for the task of grouping an arbitrary shaped database into several proper classes. Until now, a wide variety of clustering algorithms for this task have been proposed, although the majority of these algorithms are density-based. In this paper, the authors extend the dissimilarity measure to compatible measure and propose a new algorithm (ASCCN) based on the results. ASCCN is an unambiguous partition method that groups objects to compatible nucleoids, and merges these nucleoids into different clusters. The application of cluster grids significantly reduces the computational cost of ASCCN, and experiments show that ASCCN can efficiently and effectively group arbitrary shaped data points into meaningful clusters.