Journal of Frontiers of Computer Science and Technology ›› 2016, Vol. 10 ›› Issue (10): 1439-1450.DOI: 10.3778/j.issn.1673-9418.1508059

Previous Articles     Next Articles

Density Adaptive Data Competition Clustering Algorithm

SU Hui1, GE Hongwei1,2+, ZHANG Tao1   

  1. 1. School of Internet of Things Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China
    2. Ministry of Education Key Laboratory of Advanced Process Control for Light Industry, Jiangnan University, Wuxi, Jiangsu 214122, China
  • Online:2016-10-01 Published:2016-09-29

密度自适应的数据竞争聚类算法

苏  辉1,葛洪伟1,2+,张  涛1   

  1. 1. 江南大学 物联网工程学院,江苏 无锡 214122
    2. 江南大学 轻工过程先进控制教育部重点实验室,江苏 无锡 214122

Abstract: Since the existing data competition clustering algorithm has poor performance on density inhomogeneous datasets, this paper proposes a density adaptive data competition clustering algorithm. Firstly, a local density adaptive line is defined. Nextly, the density adaptive similarity can be calculated based on local density adaptive line. The density adaptive similarity can reflect the global data space distribution information and local information of data points, which can describe the relationship between data points more effectively. Then, the density adaptive similarity is used in data competition clustering algorithm. The simulation results on synthetic and real life datasets show that the proposed algorithm can obtain better performance on density inhomogeneous datasets than existing data competition clustering algorithm.

Key words: clustering, data competition, aggregation field, density inhomogeneous

摘要: 针对现有数据竞争聚类算法在处理密度不均匀数据集时聚类效果不理想的问题,提出了一种密度自适应的数据竞争聚类算法。首先,定义了一种局部密度自适应线段;然后,根据局部密度自适应线段计算出密度自适应相似度,密度自适应相似度不仅反映了数据的整体空间分布信息,还反映了数据点的局部信息,更加符合数据的实际分布;最后,将密度自适应相似度用于数据竞争聚类算法中。在人工和真实数据集上的仿真实验结果表明,新算法比现有的数据竞争聚类算法在处理密度不均匀数据集时,具有更高的聚类性能。

关键词: 聚类, 数据竞争, 聚合场, 密度不均匀