当前位置:主页 > 经济论文 > 国际贸易论文 >

基于大数据的Y公司P2P网贷业务信用风险管理平台的构建

发布时间:2018-02-01 02:10

  本文关键词: P2P网贷 信用风险管理 大数据 逻辑回归 随机森林 出处:《东华大学》2017年硕士论文 论文类型:学位论文


【摘要】:P2P网贷起源于英国,为借贷双方提供撮合平台,实现双方的借贷需求。近年来,在国家普惠金融规划的推动下,P2P发挥其普惠金融的关键作用,为小微企业和社会各阶层群体提供了更为便捷的融资和借贷渠道,极大地促进了实体经济的发展,为我国普惠金融事业做出了突出贡献。然而,由于公民信用体系尚未规范,恶意诈骗、倒闭跑路现象时有发生,整个行业积聚了巨大风险。针对以上问题,怎样更好的管理和防控借贷客户风险,已经成为一个十分突出的问题,本文在分析Y公司背景、特点以及现状的基础上,提出Y公司在信用风险管理方面存在分析方法陈旧、管理工具落后、缺乏全面性等不足问题,然后结合当前大数据分析技术,对Y公司P2P网贷业务信用风险管理平台进行构建,具有一定现实意义。研究内容主要包括:1.对国内外网贷业务风险管理的研究现状进行梳理,结合Y公司P2P网贷新信审项目,对其业务风险管理特点和策略、平台风险管理存在的问题进行分析;明确了大数据分析方法对于P2P网贷风险管理的必要性;2.建立起了适用于P2P网贷平台的借款人信用风险评估的指标体系。该体系包含反欺诈、违约预测和信用评分三个模型,通过计算用户的是否欺诈、是否违约和信用评分,来评估其信用风险的大小。并经过测试数据检验,证明了该体系具有很好的稳定性和可靠性;3.实现了大数据的管理平台的架构设计。借助于当前先进的系统开发平台和技术,构建一个大数据平台对多个业务系统数据进行统一管理,具有高可靠性、高可扩展性和高效性,并通过原型系统验证了所设计架构的可行性。本研究建立的信用风险评估的指标体系在P2P网贷行业具有可行性,能够大幅度提高风险决策效率,降低审核成本,提升风险控制水平,也必将成为未来P2P网贷企业风险管理模式的新方向。从实践意义上看,本文将大数据分析这一理论方法与Y公司P2P网贷业务风险管理背景相结合,能为Y公司新信审项目的信用风险管控提供具体的流程与步骤,其应用过程也可为国内其他P2P网贷业务的风险管控提供新的思路与决策依据,具有较高的实用价值,在一定程度上促进国内P2P网贷行业风险管理水平的提高。
[Abstract]:P2P lending originated in the United Kingdom, providing a matching platform for both lenders and borrowers to meet their lending needs. In recent years, driven by the national inclusive financial planning, P2P has played a key role in inclusive finance. For small and micro enterprises and social groups to provide a more convenient financing and lending channels, greatly promote the development of the real economy, for China's inclusive financial undertakings to make a prominent contribution. As the civil credit system has not been standardized, malicious fraud, failure to run the phenomenon has occurred from time to time, the entire industry has accumulated a huge risk. In view of the above problems, how to better manage and control the risk of lending customers. It has become a very prominent problem. Based on the analysis of the background, characteristics and current situation of Y Company, this paper puts forward that Y Company has outdated analysis methods and backward management tools in the aspect of credit risk management. Lack of comprehensiveness and other deficiencies, and then combined with the current big data analysis technology, Y company P2P network loan business credit risk management platform to build. Research content mainly includes: 1. Combing the domestic and foreign network loan business risk management research status, combined with Y company P2P network loan new credit project. The characteristics and strategies of business risk management and the problems of platform risk management are analyzed. The necessity of big data analysis method for P2P network loan risk management is clarified. 2. The index system of borrower credit risk assessment for P2P network loan platform is established. The system includes three models: anti-fraud, default prediction and credit score, and calculates whether the user is fraudulently or not. Whether default and credit score are used to evaluate the credit risk. The test data show that the system has good stability and reliability. 3. The architecture design of big data's management platform is realized. With the help of the current advanced system development platform and technology, a big data platform is constructed to manage the data of multiple business systems in a unified way, which has high reliability. High scalability and efficiency, and through the prototype system to verify the feasibility of the design framework. This study established a credit risk assessment index system in the P2P network lending industry feasibility. Can greatly improve the efficiency of risk decision-making, reduce audit costs, improve the level of risk control, but also will become the future P2P network loan enterprise risk management model of a new direction. From a practical point of view. In this paper, big data analysis of this theory and Y company P2P network loan business risk management background, can provide a specific process and steps for Y company new credit risk management project. The application process can also provide a new way of thinking and decision basis for the risk management of other P2P network loan business in China, which has a higher practical value. To a certain extent, to promote the domestic P2P network lending industry risk management level.
【学位授予单位】:东华大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:F832.4;F724.6

【参考文献】

相关期刊论文 前10条

1 程华;杨云志;曾令涛;;运用大数据挖掘传统银行业务价值:美国第一资本金融集团案例[J];南方金融;2016年08期

2 张天琪;李杨;;我国小额贷款公司发展现状、存在问题及对策[J];中国商论;2016年21期

3 高博;谢朝阳;;美国个人征信机构间的信用信息共享经验及对我国的启示[J];时代金融;2016年20期

4 保建云;;大数据金融生态系统、社会超群博弈与中国大数据金融战略[J];江苏行政学院学报;2016年04期

5 李延东;郑小娟;;大数据对解决小微企业融资难的启示——以ZestFinance公司为例[J];甘肃金融;2016年05期

6 黄莺娇;;利用互联网发展我国普惠金融的问题与建议[J];时代金融;2016年11期

7 朱良平;;基于大数据的信用风险评分模型辨析[J];中国金融电脑;2016年03期

8 于晓虹;楼文高;;基于随机森林的P2P网贷信用风险评价、预警与实证研究[J];金融理论与实践;2016年02期

9 朱友嘉;;国内P2P网络借贷平台倒闭现象原因分析[J];科技创业月刊;2016年02期

10 李胜;;大数据时代的国家社会管理现代化:模式变革与战略应对[J];广西社会科学;2016年01期

相关硕士学位论文 前4条

1 胡超;我国P2P网络借贷规范发展研究[D];浙江大学;2015年

2 张职;P2P网络借贷平台营运模式的比较、问题及对策研究[D];华东理工大学;2013年

3 郭弈;P2P网络借贷市场的融资成本与融资可获得性研究[D];西南财经大学;2011年

4 申华;基于数据挖掘的个人信用评分模型开发[D];厦门大学;2009年



本文编号:1480628

资料下载
论文发表

本文链接:https://www.wllwen.com/jingjilunwen/guojimaoyilunwen/1480628.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户c5684***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com