毕业设计(论文)论文题目:基于数据挖掘的商业银行个人信用风险评估平台设计与实现学生姓名:学生学号:专业班级:学院名称:指导老师:学院院长:2016年05月27日基于数据挖掘的商业银行个人信用风险评估平台设计与实现摘要计算机技术和信息技术的不断发展,带给我们便利的同时也带来一系列问题,其中一个问题是数据量的爆炸式增长以及数据之间的关系愈发复杂,如何对这些海量的数据进行处理,发掘隐藏在数据中潜在的理论价值和实际价值也成为生活中各个领域关注的焦点。随着我国经济的不断发展,城市和农村居民的收入和消费水平有着显著提升,个人信贷业务已经成为商业银行主营业务之一,但目前我国商业银行在个人信用风险评估方面存在着不足。因此,研究如何利用数据挖掘技术从银行现有的客户数据中分析客户的信用风险,具有重要的理论价值和实际意义。本文首先对数据挖掘的概念、发展现状进行了介绍。其次对数据挖掘的算法进行了介绍,分析了本系统会用到的数据挖掘的算法。然后,结合商业银行在客户信用风险评估方面遇到的问题进行了需求分析。在此基础上,提出了基于BP神经网络以及决策树算法的商业银行个人信用风险评估模型,为商业银行个人信用风险评估提供了可行的解决方案。关键词:数据挖掘;BP神经网络;决策树;信用风险DesignandImplementationofCommercialBanksCreditRiskAssessmentBasedonDataMiningAbstractThedevelopmentofcomputertechnologyandinformationtechnologybringusconvenience,butalsobroughtaseriesofproblems,oneoftheproblemsistherelationshipbetweentheamountofdataaswellastheexplosivegrowthofdatabetweenthemoreandmorecomplex,andhowthesemassivedataprocessingdiscoverhiddenpotentialdatatheoreticvalueandpracticalvaluehasbecomethefocusofattentioninallareasoflife.AsChina'seconomycontinuestodevelop,incomeandconsumptionlevelofurbanandruralresidentshassignificantlyimproved,consumercreditbusinesshasbecomeoneofthemainbusinessofcommercialbanks,commercialbanksinChinabutthereisalackofpersonalcreditriskassessment.Therefore,studyinghowtousedataminingtechniquestoanalyzecustomercreditriskfromthebank'sexistingcustomerdata,hasimportanttheoreticalandpracticalsignificance.Firstly,theconcept,developmentstatusdataminingareintroduced.Secondly,thedataminingalgorithmsareintroduced,weanalyzedthesystemwillusedataminingalgorithms.Then,combinedwithproblemsencounteredbycommercialbanksincustomercreditriskassessmentneedsanalysis.Onthisbasis,theproposedindividualcreditriskassessmentmodelbasedonBPneuralnetworkanddecisiontreealgorithmbasedcommercialbank,andtheproposedmodelhasbeenimprovedandvalidatedcommercialbankpersonalcreditriskprovidesafeasiblesolutionprogramevaluation.KeyWords:Datamining;BPneuralnetwork;decisiontree;CreditRisk目录第1章绪论..............................................................11.1研究背景和研究意义................................................11.1.1研究背景.....................................................11.1.2研究意义.....................................................31.2国内外研究综述....................................................41.2.1数据挖掘研究现状.............................................41.2.2商业银行信用风险研究现状.....................................61.3论文的主要工作和内容结构..........................................7第2章数据挖掘算法及相关技术............................................92.1数据挖掘的概念....................................................92.2数据挖掘的过程....................................................92.3数据挖掘的常用算法...............................................102.3.1人工神经网络................................................112.3.2决策树......................................................112.3.3遗传算法....................................................122.3.4近邻算法....................................................132.3.5k-means算法................................................132.4小结.............................................................14第3章基于BP神经网络算法的个人信用风险评估............................153.1神经网络的学习机理和机构.........................................153.1.1感知器的学习结构............................................153.1.2梯度下降法算法..............................................173.1.3反向传播(BP)算法..........................................193.2实验数据的结构及预处理...........................................223.2.1数据预处理..................................................253.2.2数据指标选取................................................263.3基于BP神经网络的信用风险评估方法模型............................293.3.1网络的构建及训练............................................293.3.2模型测试结果................................................303.4小结.............................................................33第4章基于决策树算法的个人信用风险评估.................................344.1决策树算法概述...................................................344.1.1ID3算法....................................................344.1.2C4.5算法与C5.0算法........................................374.2基于C5.0算法的决策树方法个人信用风险评估模型....................394.2.1数据采集...................................................394.2.2数据变换...................................................404.2.3决策树的构建...............................................434.2.4评估模型及模型优化.........................................484.3小结.............................................................50第5章基于数据挖掘的个人信用风险评估系统的实现.........................515.1开发环境的搭建...................................................515.1.1R语言开发环境搭建..........................................515.1.2ShinyServer安装与配置.....................................525.2模块关键功能实现.................................................535.2.1基于BP神经网络算法评估客户信用风险的实现..................545.2.2基于决策树络算法评估客户信用风险的实现.....................565.3小结.............................................................58总结与展望..............................................................59致谢....................................................................60参考文献......................................