点击次数：

论文类型： 期刊论文

发表时间： 2015-02-10

发表刊物： Journal of Information and Computational Science

收录刊物： EI、Scopus

卷号： 12

期号： 3

页面范围： 1267-1279

ISSN号： 15487741

摘要： Determination of indicators' weights is the key problem that needs to be addressed in small enterprises credit evaluation. Whether the indicators are reasonably or accurately weighted would directly affect the reliability of a credit evaluation model. Based on indicators' diversification maximized method, the paper obtains combination weighting model from three single weighting models including G1 model, CRITIC model and difficulty weighting model, establishes small enterprises credit risk evaluation model and tests the evaluation results by using different weighting models. The features and innovations of this paper are: first, based on the idea of deviation of indicators' value from indicators' average value maximizing, it determines the coefficient of combination weight by integrating the subjective and objective weights. It avoids the problem lying in the existing study that overlooks the diversification of indicator's value between overall statuses of indicators. Second, the evaluation results derived from different weighting models pass Spearman rank correlation coefficient test. It means that the credit evaluation result based on combination weighting model is in accordance with results from three single weighting models. Third, result of empirical study demonstrates that evaluation result based on indicators' diversification maximized combination weighting model achieves the best consistency and lowest FRR and FAR. ?, 2015, Binary Information Press. All right reserved.