P2P网贷个人信用风险评估模型研究——基于混合果蝇神经网络的方法
发布时间：2019-01-01 15:48
【摘要】：P2P网贷在爆发式增长的同时,也面临着重大的信用风险,个人信用评估是降低信用风险的重要方法。根据P2P网贷自身的特点,对影响P2P网贷借款人信用风险的因素进行分析,引入互联网信息领域特有的风险因素,建立了P2P网贷个人信用风险评估指标体系。基于该指标体系,考虑P2P网贷中"软信息"较多、"硬信息"缺失的特点,提出了基于BP神经网络的信用评估模型。为了提高BP神经网络的收敛速度和精度,将改进的果蝇优化算法作为BP神经网络的学习算法,对神经网络的权重进行训练。通过"人人贷"平台收集的样本数据进行实验验证。结果表明:改进果蝇神经网络评估模型比传统BP神经网络模型有更强的学习能力和预测能力,是P2P网贷个人信用风险评估的有效方法。
[Abstract]:P2P network loan is facing great credit risk while it is exploding. Personal credit evaluation is an important method to reduce credit risk. According to the characteristics of P2P network loan, this paper analyzes the factors that affect the credit risk of P2P network loan borrowers, introduces the unique risk factors in the field of Internet information, and establishes the evaluation index system of P2P network loan individual credit risk. Based on this index system, considering the characteristics of "soft information" and "hard information" in P2P network loan, a credit evaluation model based on BP neural network is proposed. In order to improve the convergence speed and accuracy of the BP neural network, the improved Drosophila optimization algorithm is used as the learning algorithm of the BP neural network, and the weight of the neural network is trained. Through the "everyone loan" platform collected sample data for experimental verification. The results show that the improved evaluation model of Drosophila neural network has stronger learning ability and prediction ability than the traditional BP neural network model, and it is an effective method to evaluate the credit risk of individuals in P2P network.
【作者单位】： 南京工业大学经济与管理学院;
【基金】：国家自然科学基金面上项目(71371097) 南京工业大学科研项目(ZKJ201531)
【分类号】：F724.6;F832.4
[Abstract]:P2P network loan is facing great credit risk while it is exploding. Personal credit evaluation is an important method to reduce credit risk. According to the characteristics of P2P network loan, this paper analyzes the factors that affect the credit risk of P2P network loan borrowers, introduces the unique risk factors in the field of Internet information, and establishes the evaluation index system of P2P network loan individual credit risk. Based on this index system, considering the characteristics of "soft information" and "hard information" in P2P network loan, a credit evaluation model based on BP neural network is proposed. In order to improve the convergence speed and accuracy of the BP neural network, the improved Drosophila optimization algorithm is used as the learning algorithm of the BP neural network, and the weight of the neural network is trained. Through the "everyone loan" platform collected sample data for experimental verification. The results show that the improved evaluation model of Drosophila neural network has stronger learning ability and prediction ability than the traditional BP neural network model, and it is an effective method to evaluate the credit risk of individuals in P2P network.
【作者单位】： 南京工业大学经济与管理学院;
【基金】：国家自然科学基金面上项目(71371097) 南京工业大学科研项目(ZKJ201531)
【分类号】：F724.6;F832.4
【相似文献】
相关期刊论文 前10条
1 杨硕;许可;黄鹏;;基于SOM神经网络技术的卷烟零售客户自组织分类[J];商场现代化;2009年04期
2 朱岩;;数字神经向大智慧延伸——扫描神经网络技术的商业应用[J];互联网周刊;1999年17期
3 张世浩;;精准广告 再造P2P时代网络盈利新模式[J];中国广告;2006年10期
4 倪月娟;;我国P2P小额网络信贷研究[J];现代商业;2013年27期
5 余凡;余红伟;许伟;;基于BP神经网络的消费者网络质量安全信息预警实证研究[J];宏观质量研究;2014年01期
6 刘增明;陈运非;蒋海青;;基于PCA-BP神经网络方法的供应商选择[J];工业工程与管理;2014年01期
7 ;P2P无线东京大登陆[J];网际商务;2001年07期
8 张志强;;当前我国P2P信贷现状及发展对策[J];华北金融;2013年07期
9 樊融杰;;P2P入侵外汇交易[J];英才;2014年03期
10 朱s，
本文编号：2397768
本文链接：https://www.wllwen.com/jingjilunwen/touziyanjiulunwen/2397768.html
