犹豫模糊数据对象集的谱聚类算法  

Spectral Clustering Algorithm for Hesitating Fuzzy Data Object Set

作　　者：孙爽爽 黄德才[1] 陆亿红[1] SUN Shuang-shuang;HUANG De-cai;LU Yi-hong(College of Computer Science and Technology,Zhejiang University of Techonology,Hangzhou 310023,China)

机构地区：[1]浙江工业大学计算机科学与技术学院,杭州310023

出　　处：《小型微型计算机系统》2023年第2期225-231,共7页Journal of Chinese Computer Systems

基　　金：浙江省基础公益研究计划项目(LGG19E090001)资助.

摘　　要：目前适用于犹豫模糊数据对象集的聚类算法研究仍然非常有限,现有的犹豫模糊数据对象集层次聚类算法受异常点影响较大且容易聚成链状.针对上述问题,本文首先提出了一种可扩展的犹豫模糊集的加权相似度计算方法,该方法不仅可以利用不同的函数计算相似度,而且可以根据实际问题构造最优的相似度函数.基于该加权相似度计算方法,结合经典的谱聚类算法提出了犹豫模糊数据对象集的谱聚类算法(SCHF).针对目前国内外还没有可用于犹豫模糊数据对象集聚类的标准数据集的现实情况,本文提出了一种确定性数据的犹豫模糊方法并在仿真实验中应用.仿真实验不仅验证了SCHF算法的有效性,而且表明SCHF算法比两种已知算法有更好的聚类效果.At present,the research on clustering algorithms for hesitating fuzzy data object sets is still very limited.The existing hierarchical clustering algorithms for hesitating fuzzy data object sets are greatly affected by abnormal points and are easy to cluster into chains.In response to the above problems,this paper first proposes a scalable weighted similarity calculation method for hesitant fuzzy sets.This method can not only use different functions to calculate the similarity,but also construct the optimal similarity function according to the actual problem.Based on the weighted similarity calculation method,combined with the classic spectral clustering algorithm,a spectral clustering algorithm for hesitating fuzzy data object sets(SCHF)is proposed.In view of the reality that there is no standard data set for clustering hesitation fuzzy data object sets at home and abroad,this paper proposes a hesitation fuzzy method for deterministic data and applies it in simulation experiments.Simulation experiments not only verify the effectiveness of the SCHF algorithm,but also show that the SCHF algorithm has a better clustering effect than the two known algorithms.

关 键 词：犹豫模糊数据对象集 相似度 谱聚类 数据挖掘 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]