基于横向联邦学习的小微企业信贷决策模型研究  

Research on credit decision model of small and micro enterprises based on horizontal federal learning

作　　者：朱军[1] ZHU Jun(Guizhou Provincial Branch of the People’s Bank of China,Guiyang 550001,China)

机构地区：[1]中国人民银行贵州省分行,贵阳550001

出　　处：《智能计算机与应用》2024年第2期97-99,共3页Intelligent Computer and Applications

摘　　要：本文对已有影响信贷决策的因素进行研究,综合得出影响银行信贷决策的指标,在使用PCA对指标数据进行简化的基础上,运用横向联邦学习算法,实现不同金融机构间信贷数据共享,得出精准信贷决策模型和小微企业贷款额度,这对解决小微企业融资难、融资慢以及银行担忧信贷风险而对小微企业惜贷的问题有重大意义。This paper studies the factors influencing credit decisions and synthesizes indicators that affect bank lending decisions.Upon simplifying the indicator data using PCA(Principal Component Analysis),it employs a horizontal federated learning algorithm to facilitate the sharing of credit data among different financial institutions.This approach results in an accurate credit decision model and loan quotas for small and micro enterprises.This is significantly meaningful in addressing the difficulties small and micro enterprises face in financing,such as slow funding processes and banks′reluctance to lend due to concerns over credit risks.

关 键 词：PCA 横向联邦学习 数据共享 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]