联系我们

论文查询结果	返回搜索

论文编号：	11018
作者编号：	2120142584
上传时间：	2019/6/17 15:07:50
中文题目：	消费金融授信管理研究 ——以M银行信用卡中心为例
英文题目：	Research on Audit Management of Consumer Finance -A case study of M Bank Credit Card Center
指导老师：	苏东海
中文关键字：	消费金融；大数据风控；风险；授信管理
英文关键字：	Consumer Finance；Bigdata Risk control；Risk；Audit management
中文摘要：	目前，随着消费金融行业的快速发展和大数据风控技术不断完善，越来越多的企业发力开拓消费金融业务。众所周知，消费金融业务能否做好、做大，风险管理是其核心之一。在业务发展的初期，授信管理相比于贷中、贷后管理更为重要。授信管理是产品设计、进件渠道管理、系统建设、风险建模、授信流程和额度评估的统一结合体，并不能孤立的认为授信管理仅仅是风险规则的简单制定。在众多企业纷纷踏足此领域时，客户体验与盈利能力便是竞争的重要筹码。只有结合人工智能在现有技术水平下做好大数据风控并建立强大的决策引擎系统以及智能报表指标系统和电核系统，让更少人工干预，才能打造出纯线上大数据风控体系,这也必然是大势所趋。 本文介绍了风险评分卡的建模流程以及信用卡中心的授信管理，创新性的总结了行业内存在的诸多问题，并以人工智能为导向，提出了较新颖的优化方案。本文首先分析了目前的市场环境，介绍了消费金融行业现有的参与主体、产品模式、系统建设模式、风险建模模式和授信模式，并重点介绍了风险评分卡开发标准流程，并以M银行信用卡中心大数据风控的授信流程、反欺诈规则和信用政策、人工电核和额度评估模型为行业代表案例研究对象，指出了其授信管理的五个不足之处，即组织架构不合理、报表指标体系不完善、决策引擎系统不灵活、风险建模人员不足和人工电核不足。 针对授信管理的不足，本文提供了相应的优化方案。授信管理应形成“自上而下”和“自下而上”相结合的管理机制，针对组织架构不合理的不足提供了建设海军陆战队似的风险管理架构；针对报表指标体系，提供了智能关联分析的报表指标体系设计想法；针对决策引擎，提供了开发流程配置与规则配置一体的决策引擎系统概念；针对风险建模人员的培养方面，提供了轮岗、培训和联合建模的方案；针对人工电核，提供了智能电核的初级与高级方案。
英文摘要：	At present, with the rapid development of the consumer finance industry and the continuous improvement of the big data wind control technology, more and more enterprise force develops the consumption finance business. It is well known that the consumption financial business can be well done, and the risk management is one of its core. In the early stage of business development, audit management is more important than in loan and post-loan management. Audit management is a unified combination of product design, channel management, IT construction, risk modeling, audit process and credit line evaluation. It is not considered that audit management is only a simple formulation of risk rules. When a large number of enterprises are in the field, the customer experience and profitability is an important chip for competition. It is inevitable that the big data risk control system on the pure line can be produced only by combining artificial intelligence with the existing technical level, and establishing a strong decision-making engine system and the intelligent BI system and the electric nuclear system, so that the big data risk control system on the pure line can be produced. This paper introduces the modeling process of the risk score card and the audit management of the credit card center, and summarizes the problems existing in the industry, and puts forward a novel optimization scheme based on artificial intelligence. This paper first analyzes the present market environment, introduces the existing participating main body, the product model, the IT construction mode, the risk modeling mode and the audit mode of the consumer finance industry, and mainly introduces the development standard process of the risk score card. The paper points out the five shortcomings of the audit management, that is, the organization structure is not reasonable and the BI system is not perfect, the decision-making engine system is not flexible, the risk modeling personnel are not enough, and the artificial electric core is not sufficient. In view of the shortage of audit management, this paper provides the corresponding optimization scheme. The audit management should form the management mechanism of the combination of the "Top-down" and the "Bottom-up", and provides a marine-like risk management framework for the unreasonable shortage of the organization structure; for the BI system, the BI design idea of the intelligent association analysis is provided; and aiming at the decision engine, The concept of decision-making engine system that integrates the development process configuration and the rule configuration is provided. In view of the training of risk modeling personnel, a scheme of wheel-job, training and joint modeling is provided, and the primary and high-level scheme of the intelligent electric core is provided for the artificial electric core.
查看全文：	预览  下载(下载需要进行登录)