公平谱聚类方法用于提高簇的公平性  

Fair Method for Spectral Clustering to Improve Intra-cluster Fairness

作　　者：徐夏 张晖[1] 杨春明[1] 李波[1] 赵旭剑[1] XU Xia;ZHANG Hui;YANG Chunming;LI Bo;ZHAO Xujian(School of Computer Science and Technology,Southwest University of Science and Technology,Mianyang,Sichuan 621010,China)

机构地区：[1]西南科技大学计算机科学与技术学院,四川绵阳621010

出　　处：《计算机科学》2023年第2期158-165,共8页Computer Science

基　　金：四川省科技厅重点研发项目(2021YFG0031);四川省省级科研院所科技成果转化项目(2022JDZH0035)。

摘　　要：最近,算法的公平性问题引起了机器学习领域学者的广泛讨论。鉴于谱聚类在现代数据科学中的广泛流行,研究谱聚类的算法公平性是一个至关重要的话题。现有的公平谱聚类算法主要存在两个缺点:1)公平性能差;2)仅在单个敏感属性下工作。文中将公平问题视为一种约束谱聚类问题,通过求解约束谱聚类的可行解集,提出了一种非规范化公平谱聚类方法(Unnormalized Fair Spectral Clustering,UFSC),用于提升公平性能。此外,文中还提出了一种适用于多个敏感属性约束的公平聚类算法(Multi-sensitive Attributes Fair Spectral Clustering,MFSC)。在多个真实数据集上进行了实验,结果表明,UFSC和MFSC算法比现有的公平谱聚类算法生成的聚类结果更加公平。Recently,the fairness of the algorithm has aroused extensive discussion in the machine learning community.Given the widespread popularity of spectral clustering in modern data science,studying the algorithm fairness of spectral clustering is a crucial topic.Existing fair spectral clustering algorithms have two shortcomings:1)poor fairness performance;2)work only for single sensitive attribute.In this paper,the fair spectral clustering problem is regarded as a constrained spectral clustering problem.By solving the feasible solution set of constrained spectral clustering,an unnormalized fair spectral clustering(UFSC)method is proposed to improve fairness performance.In addition,the paper also proposes a fair clustering algorithm suitable for multiple sensitive attribute constraints.Experimental results on multiple real-world datasets demonstrate that the UFSC and MFSC are fairer than the existing fair spectral clustering algorithms.

关 键 词：算法公平性 公平谱聚类 约束谱聚类 机器学习 数据分析 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]