A Survey on clustering Current status and challenging issues
A Survey on clustering Current status and challenging issues
复制标题
集群现状及挑战问题综述
DOI：
--
复制
发表时间：
2010
期刊：
影响因子：
--
通讯作者：
Salim Jiwani
中科院分区：
文献类型：
--
作者：
Salim Jiwani
- 聚类是从数据集中提取子集的艺术。它有助于识别隐藏的信息，并根据属性或属性集将数据排列到其逻辑组中。本文的目的是探讨各种聚类方法，并简要介绍他们的工作方式，使研究人员可以有一个局部的看法所讨论的方法。这项工作提出了收益和陷阱，也在一定程度上相关的最坏情况下的复杂性，每种聚类方法。这里讨论的技术只是聚类算法的一个快照。目前，基于模型的算法被用来提高聚类算法的效率。本文将有助于设计的算法选择这样一个目的。
— Clustering is the art of subset from a dataset. It helps in identifying the hidden information and arranging data into its logical group based on an attribute or a set of attributes. The intent of this paper is to explore a variety of clustering methods and brief their working styles so that researches can have a partial view of methods discussed. This work presents gains and pitfalls and also associated worst case complexities of each clustering method to some extent. The techniques discussed here are just a snap shot of clustering algorithms. Currently model based algorithms are been used to improve efficiency of clustering algorithms. This paper would be helpful in devising the choice of algorithm for such a purpose.