融合SOM神经网络与K-means聚类算法的用户信用画像研究
User credit profile integrating SOM Neural network and K-means clustering algorithm
罗博炜 1罗万红 2谭家驹2
作者信息
- 1. 深圳大学 数学科学学院,深圳 518060
- 2. 五邑大学 数学与计算科学学院,江门 529030
- 折叠
摘要
为提高现阶段基于K-Means聚类算法的用户信用画像模型的准确性和实时性,提出一种融合自组织映射(SOM,Self-Organizing Map)神经网络与K-Means聚类算法的改进方法.通过SOM对用户数据进行降维和特征提取,直接获得最优聚类数目后再用K-Means算法进行聚类分析.通过真实在线借贷平台数据对所提方法进行验证,结果表明,该方法可提升用户信用画像分析的质量,更好地满足金融数据分析中对实时管理和风险控制的要求,为金融机构提供精准的决策支持.
Abstract
To improve the accuracy and real-time performance of user credit profile models based on K-Means clustering algorithm,this paper proposed an improved method that integrated Self Organizing Map(SOM)neural network with K-Means clustering algorithm.The paper used SOM to reduce dimensionality and extract features from user data,directly obtained the optimal number of clusters,and then used K-Means algorithm for clustering analysis,validated the proposed method through real online lending platform data.The results show that the proposed method can improve the quality of user credit profile analysis,better meet the requirements of real-time management and risk control in financial data analysis,and provide accurate decision support for financial institutions.
关键词
用户信用画像/SOM神经网络/K-means聚类算法/时间复杂度/风险控制Key words
user credit profile/SOM(Self-Organizing Map)neural network/K-Means clustering algorithm/time complexity/risk control引用本文复制引用
基金项目
国家自然科学基金青年项目(62101388)
广东省高等教育教学改革项目(GDJX2020016)
出版年
2024
NETL
NSTL
万方数据