基于BP神经网络模型的我国寿险公司信用评级研究

发布时间：2018-03-06 05:27

本文选题：寿险公司　切入点：信用评级　出处：《哈尔滨工程大学》2013年硕士论文　论文类型：学位论文

【摘要】：诚信作为寿险公司存在的基础，是寿险公司健康发展的前提，也是寿险公司提升竞争力和建立良好社会形象的重要因素。但是，由于消费者与寿险公司之间的信息不对称性，造成了消费者对寿险公司的低信任度，使寿险公司遭受诚信危机。信用评级的产生源自信息不对称理论。信用评级是对企业内在质量的全面检验和考核，从而能够对寿险公司的偿付能力进行综合评价，是专业性很强的有针对性的信用分析。对寿险公司进行信用评级可以为消费者提供准确的风险信息，使消费者掌握寿险公司的真实情况，从而增加消费者对寿险公司的信任度，促进寿险行业的发展。本文首先对国内外相关的文献进行了回顾和总结，从信用评级理论、信息不对称理论、全面风险管理思想以及信用要素分析法对我国寿险公司信用评级进行了理论分析，为后文的研究提供了理论支持。其次，通过学习国外寿险公司信用评级指标体系构建方法，并考虑中国寿险公司的自身特点，建立我国寿险公司信用评级指标体系，并进行了适用性分析，为下文进一步研究奠定了基础。再次，，对传统的BP神经网络模型进行了简单分析，然后论述了在此基础上发展起来的改进的BP神经网络模型。本文运用改进的BP神经网络模型，利用十五家寿险公司样本的各指标值和实际评级结果，运用Matlab软件进行仿真训练，使网络模型输出的评级结果与实际评级结果之间的总体误差满足设定要求，从而训练出寿险公司信用评级模型。最后，通过一个实例来验证训练的网络模型对寿险公司进行信用评级的准确性和适用性，结果表明训练的网络模型得出的评级结果符合该公司的实际情况，比较正确地反映了公司的信用等级。因此，可以通过此模型对寿险公司的信用等级进行预测。
[Abstract]:Honesty, as the foundation of life insurance company, is the premise for life insurance company to develop healthily, and it is also an important factor for life insurance company to improve its competitiveness and establish a good social image. However, due to the asymmetry of information between consumers and life insurance companies, It causes consumers' low trust in life insurance companies and makes life insurance companies suffer from credit crisis. Credit rating comes from information asymmetry theory. Credit rating is a comprehensive test and assessment of the internal quality of an enterprise. Therefore, the comprehensive evaluation of the solvency of life insurance companies is a highly professional and targeted credit analysis. The credit rating of life insurance companies can provide consumers with accurate risk information. To make consumers master the true situation of life insurance companies, thereby increasing consumers' trust in life insurance companies and promoting the development of life insurance industry. Firstly, this paper reviews and summarizes the relevant literatures at home and abroad, and analyzes the credit rating of Chinese life insurance companies theoretically from the credit rating theory, the information asymmetry theory, the overall risk management thought and the credit factor analysis method. Secondly, by studying the construction method of foreign life insurance company credit rating index system and considering the characteristics of Chinese life insurance company, we establish our country life insurance company credit rating index system. The applicability analysis is carried out, which lays a foundation for further research. Thirdly, the traditional BP neural network model is simply analyzed. Then it discusses the improved BP neural network model, which is developed on this basis. This paper uses the improved BP neural network model, using the index value of 15 life insurance company samples and the actual rating result, and using Matlab software to carry on the simulation training. Make the overall error between the network model output and the actual rating results meet the set requirements, thus train the life insurance company credit rating model. Finally, An example is given to verify the accuracy and applicability of the training network model for life insurance company credit rating. The results show that the network model obtained by the training model is in line with the actual situation of the company. Therefore, this model can be used to predict the credit rating of life insurance companies.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F842.3;TP18

【参考文献】