一种基于Divide-and-Merge聚类算法的改进算法

An Improved Algorithm Based on Divide-and-Merge Clustering Algorithm

BNAK-Divide-and-Merge聚类算法是基于David等人提出的Divide-and-Merge算法的一种改进算法。Divide-and-Merge算法是一种将自顶向下的分裂方法和自底向上的聚合方法相结合的聚类算法。虽然这个聚类算法已经通过众多实验表明其聚类的效率和质量,但是它在数据集很大的情况下分裂会很耗时间和空间资源,并且它需要阈值来确定聚类个数的方法也不是很理想。针对以上两个主要不足,对原算法进行改进。

BNAK-Divide-and-Merge clustering algorithm is an improved algorithm which is based.on the Divide-and-Merge clustering algorithm proposed by David et al. Divide-and-Merge is a methodology which combines a top-down divide method with a bottom-up merge method. Although it has been proved to be a method with high efficiency and quality of clustering by implementing lots of relevant experiment, its divide phase will consume too much time and space resources when it is applied to very huge sets; furthermore the method which can figure out the number of clustering with a threshold is also not best. Accordingly, improves the original algorithm to overcome the two major flaws mentioned above.