...
首页> 外文期刊>Information Technology Journal >Research of Clustering Evaluation Index for User Grouping
【24h】

Research of Clustering Evaluation Index for User Grouping

机译:用户分组聚类评价指标研究

获取原文
           

摘要

Some research had devoted to users grouping based on the clustering method. However, in many cases, the distribution of user?s character is unknown, so that it is difficult to decide which algorithm is more suitable for user grouping. In this study an improved clustering evaluation index named as SSDS (inner-cluster Scattering, extra-cluster Separation, Density of centroids, balance of Size), was proposed. SSDS is able to select suitable algorithm for a data set and obtain the optimal clustering scheme though balancing among the weight of density, the degree of scattering and the size of clusters. Results of experiment for learner grouping indicated that this evaluation index method is feasible and efficient. In this experiment, the FarthestFirst algorithm with number of clusters set as 5 is the best clustering scheme.
机译:一些研究致力于基于聚类方法的用户分组。但是,在许多情况下,用户角色的分布是未知的,因此很难确定哪种算法更适合用户分组。在这项研究中,提出了一种改进的聚类评估指标,称为SSDS(内部集群散射,集群外分离,质心密度,大小平衡)。通过在密度权重,分散程度和簇大小之间取得平衡,SSDS能够为数据集选择合适的算法并获得最佳的簇方案。学习者分组的实验结果表明,该评价指标方法是可行和有效的。在本实验中,将群集数设置为5的FarthestFirst算法是最佳的群集方案。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号