首先,本文在对山东省JH小额贷款等公司进行实地调研基础上,选
取2013年-2015年数据为样本,结合专家访谈、文献参考等方式,针对小
贷公司经营特点分别建立适合个人客户和公司客户的信用风险指标体系,
并运用正态检验、非参数检验以及因子分析进行指标筛选;其次,根据不
同客户群体指标体系及样本数据的特点,选用CHAID算法作为个人客户
信用风险预警的方法,选用C5.0算法作为公司客户信用风险预警的方法,
通过训练样本构建决策树模型,运用检验样本进行模型的准确性验证
实证研宄结果表明,小额贷款公司的信用风险影响因素复杂,需要根
据不同客户群体构建指标体系,并选取进入模型的指标变量;针对个人客
户与公司客户选择适合的算法构建基于决策树技术的小贷公司信用风险
预警模型,可以较为准确的预测信用风险的发生情况,具有很强的实用性,
有利于小贷公司进行风险防范,降低违约率,促进行业的持续健康发展
关键词:小额贷款公司,信用风险预警,决策树技术
I
ABSTRACT
RESEARCH ON CREDIT RISK EARLY WARNING OF
SMALL LOAN COMPANY BASED ON DECISION TREE
TECHNOLOGY
ABSTRACT
Since 2014, the national economy is facing downward pressure, part of
the enterprises have the phenomenon of enterprise insolvency, capital chain
rupture and the debt is not returned, etc. These make microfinance companies,
which focus on lend as the main business, face great challenges of survival.
Therefore, according to the operating characteristics of small loan companies,
we build a scientific and reasonable credit risk early warning model. It is of
great theoretical and practical significance to use the model to predict the
default risk and reduce the probability of risk.
First of all, on the basis of field investigation of the JH small loan
company in Shandong province, this paper selects data from 2013 to 2015 as
samples. Combined with expert interviews and literature reference method,
this paper established credit risk index system for individual customers and
corporate clients according to the operating characteristics of small loan
companies. By using normal test, nonparametric test and factor analysis
method to carry on the index selection. Secondly, according to the difference
of the index system and the characteristics of the sample data for individual
customers and corporate clients, we use the CHAID algorithm as the method
of individual customer credit risk early warning, and the C5.0 algorithm as the
method of company customer credit risk early warning. In this paper, the
decision tree model is constructed by training samples, and the accuracy of the
model is verified by the test samples.
in
北京化工大学硕士学位论文
The empirical results show that the small loan company&39;s credit risk
factors are complex. Therefore, it is necessary to build the index system
according to the different customer groups, and select the index variables into
the model. According to the characteristics of individual customers and
corporate customers to select a suitable algorithm to construct a credit risk
early-warning model of microfinance companies based on decision tree
technology, which can accurately predict credit risk. The early warning model
of credit risk has strong practicability, which is conducive to the small loan
companies to prevent risks, reduce the default rate and promote the sustained
and healthy development of the industry.
KEY WORDS: Small Loan Company, Credit Risk Early Warning, Decision
Tree Technology
IV
目录
第一韋mt i
i.i研宄背景及意义l.i.i研宄背景 1
1.1.2研究目的 2
1.1.3研宄意义 2
1.2研宄内容和方法1.2.1研宄内容 3
1.2.2研宄方法 6
1.3本文创新点 6
第二章文献鄉与理论基础2.1文献综述 7
2.1.1国外小额信贷的研宄现状2.1.2国内小额贷款公司的研究成果2.1.3研究述评 13
2.2小额贷款公司信用风险预警的理论分析2.2.1信用风险悖论2.2.2软信息和关系型借贷理论2.2.3信息不对称理论第三章小额贷款公司信用风险指标体系的设计3.1小额贷款公司信用风险影响因素分析3.2信用风险预警指标的选取原则3.3信用风险预警指标体系的建立3.3.1原始指标的选择3.3.2样本的选择
22
3.3.3指标的筛选
22
第四章基于决策树技术的小贷公司信用风险预警实证研究
29
V
北京化工大学硕士学位论文
4.1小贷公司信用风险预警模型的选择
29
4.1.1信用风险预警模型的比较分析
29
4.1.2决策树模型的特点
30
4.2基于决策树技术的个人客户信用风险预警
31
4.2.1个人客户信用风险预警决策树算法的选取
31
4.2.2模型参数设置
31
4.2.3模型的构建及输出
32
4.2.4个人客户信用风险预警结果分析
34
4.3基于决策树技术的公司客户信用风险预警
35
4.3.1公司客户信用风险预警决策树算法的选取
35
4.3.2 C 5.0算法最佳分组变量的确定
35
4.3.3模型参数设置
36
4.3.4模型的构建及输出
37
4.3.5公司客户信用风险预警结果分析
44
第五章结论与建议
47
5.1研宄结论 47
5.2相关建议 48
5.2.]加强信用风险控制的外部管理措施
48
5.2.2加强信用风险控制的内部管理机制
49
文献 51
附录 55
麟 71
研究成果及发表的学术论文
73
导师和作者简介
75
VI
CONTENTS
CONTENTS
Chapter 1 introduction1.1 Research Background and Significance1.1.1 Research Background1.1.2 Research Objective1.1.3 Research Meaning1.2 Research Contents and Methods1.2.1 Research Contents1.2.2 Research Methods1.3 Innovation 6
Chapter 2 Literature Review and Theoretical Basis2.1 Literature Review2.1.1 Research Status of Foreign Microfinance2.1.2 Research Results of Domestic Small Loan Company2.1.3 Research Review2.2 Theoretical Analysis of Credit Risk Early Warning for Small Loan Company2.2.1 Paradox of Credit Risk2.2.2 Soft Information and Relationship Lending Theory2.2.3 Asymmetric Information TheoryChapter 3 Design of Credit Risk Index System of Small Loan Company....3.1 Analysis of Credit Risk Factors of Small Loan Company3.2 Credit Risk Early Warning Index Selection Principle3.3 Establishment of Credit Risk Early Warning Index System3.3.1 Selection of Original Index3.3.2 Sample Selection
22
3.3.3 Index Screening
22
VII
北京化工大学硕士学位论文
Chapter 4 Empirical Research on Credit Risk Early Warning of Small
loan company based on Decision Tree Technology
29
4.1 Selection of Credit Risk Early Warning Model for Small Loan Company
29
4.1.1 Comparative Analysis of Credit Risk Early Warning Model
29
4.1.2 Characteristics of Decision Tree Model
30
4.2 Credit Risk Early warning of Individual Customer Based on Decision Tree
31
4.2.1 Selection of Decision Tree Algorithm
31
4.2.2 Model Parameter Setting
31
4.2.3 Model Construction and Output
32
4.2.4 Result Analysis
34
4.3 Credit Risk Early warning of Company Customer Based on Decision Tree
35
4.3.1 Selection of Decision Tree Algorithm
35
4.3.2 Determination of Optimal Grouping Variables in C 5.0 algorithm
35
4.3.3 Model Parameter Setting
36
4.3.4 Model Construction and Output
37
4.3.5 Result Analysis
44
Chapter 5 Conclusion and Suggestion
47
5.1
。。。以上简介无排版格式，详细内容请下载查看
==>>