现代信息科技2025，Vol.9Issue(21)：122-126,5.DOI:10.19850/j.cnki.2096-4706.2025.21.023
多维属性融合构建亲和度的谱聚类分析算法研究
Research on Spectral Clustering Analysis Algorithm of Constructing Affinity by Fusing Multi-dimensional Attributes
摘要
Abstract
In the context of the deep integration of information technologies and social networks,multi-dimensional data has shown explosive growth,and its characteristics of high dimensionality,complex structure and hidden correlation have become the core problems in the field of data analysis.In the face of such problems,traditional clustering methods often have poor clustering effect due to problems such as dimensional redundancy,significant sparsity,and single similarity measurement,which are difficult to meet the needs of actual scenarios.This study aims to extract several independent factors with high correlation and weak correlation between factors based on factor analysis method to reduce the dimensionality of multi-dimensional data,select appropriate affinity algorithms for different types of data,construct a composite affinity matrix with dynamic weights and fusion of multiple affinity algorithms,and generate an affinity matrix suitable for specific application scenarios,which is finally used for spectral clustering analysis.This method can not only improve the clustering accuracy,but also provide more accurate data support for application scenarios such as personalized recommendation and public opinion analysis.关键词
因子分析/谱聚类/亲和度/多维数据/动态权值Key words
factor analysis/spectral clustering/affinity/multi-dimensional data/dynamic weight分类
计算机与自动化引用本文复制引用
杨婧婷,王湘榆,马宏丹,梁瑞文,葛东旭..多维属性融合构建亲和度的谱聚类分析算法研究[J].现代信息科技,2025,9(21):122-126,5.基金项目
南京审计大学金审学院2024年江苏省大学生实践创新训练计划项目(202413994049Y) （202413994049Y）
