论文部分内容阅读
研究了信用分析中的软计算方法——BP网络与自适应神经模糊推理,并对软计算方法与DEA、TOPSIS的结合运用作了进一步探讨。采用V-foldCross-validation技巧,利用样本公司实际指标数据对该模型的评分效果进行了实证研究。
The soft computing method in credit analysis - BP network and adaptive neuro-fuzzy inference is studied. The combination of soft computing method with DEA and TOPSIS is discussed further. The V-foldCross-validation technique was used to make an empirical study on the scoring effect of the model using the actual data of the sample companies.