中文摘要：

摘要：传统的模糊评估方法对于信用平稳下的信用风险测度具有一定的可行性，但是在信用突变下，模糊评估方法存在功能局限，容易引发测度等级的过度跳跃。对此，本文基于偏好信息熵与物元可拓理论相结合的偏好熵权物元可拓方法，构建了信用突变下的商业银行信用风险测度模型，并以苏州地区某化工企业为样本，选取其在不同时间点的实测数据，对测度模型进行了实证分析。研究表明，宏观信用环境的突变，促使样本企业的信用风险水平出现了小幅度提高，但尚未引发其信用风险等级在短期内出现“跳跃式”突变，其信用风险等级始终处于轻度状态。这足以表明，偏好熵权物元可拓模型可以很好地解决信用突变下的商业银行信用风险测度难题。该研究成果将为我国商业银行体系构建科学高效的信用风险管控机制，提供重要的理论指导与决策参考。

英文摘要：

Traditional credit risk model based on fuzzy assessment technique is useful when credit is stable but has limitations under credit mutation as it tends to over-measure credit ratings with jumps. This paper constructs a commercial bank credit risk model based on preference entropy-weight and matter-element extension techniques. This model is then applied to measure commercial bank credit risk with real data from some sample chemistry enterprises of different periods in Suzhou. The model predicts that, macro-credit environment mutation causes credit risk ratings of sample chemistry enterprise to drop slightly, but it never makes credit risk ratings present a ＂jump＂ mutation in the short term. In reality, credit risk ratings of these sample chemistry enterprise always maintains at high grade, which shows that commercial bank credit risk measurement model based on entropy-weight and matter-element extension is better in credit risk measurement under credit mutation.