基于模糊神经网络的小微企业信用评级研究

当前小微企业贷款需求日益增加,建立行之有效的小微企业信用评级模型已成为学术界和实务界关注的焦点.本文在阐述模型和构建指标体系的基础上,提出基于模糊神经网络开展小微企业信用评级的研究步骤,以某农村商业银行小微企业信贷微观数据为实证样本,分别进行小型企业和微型企业信用评级检测.实证结果表明,模糊神经网络模型在小微企业信用评级研究中具有较BP神经网络模型更高的检测精度.模型能够实现评级主观性与客观性结合,可对数据进行定性调节和批量处理,且具有明确的计算过程和决策规则,故适用于信用评级研究且具有稳健性.

Research on the credit rating of small and micro enterprises based on fuzzy neural network

Currently, the loan demand of small and micro enterprise is rapidly increasing, thus the establishment of an effective credit rating model for small and micro enterprises has become the focus of attention in the academic and practical fields.On the basis of the model and the index system, this paper puts forward the research steps of small and micro enterprises credit rating based on fuzzy neural network.This paper takes the micro data of the small and micro enterprises in a rural commercial bank as the empirical sample, and carry out the small enterprises and micro enterprises credit rating test separately.The empirical results show that the fuzzy neural network model in the small and micro enterprises credit rating research brought a higher detection accuracy than the BP neural network model.The model can realize the combination of subjective and objective ratings;can be used for qualitative adjustment and batch processing of data;and has explicit calculation process and decision rules.Therefore, it is suitable for the research of credit rating and has robustness.