作　　者： ; ;

机构地区： 五邑大学系统科学与技术研究所

出　　处： 《辽宁工程技术大学学报（自然科学版）》 2006年第B06期254-256,共3页

摘　　要： 针对商业银行的行业特点以及其中存在的信贷风险问题,指出:反映商业银行运营状况的各种指标数据存在高度的非线性、不确定性和不精确性,并由此导致现有的信贷风险评估模型难以胜任。为此,结合统计学习理论的研究成果,建立了基于最小一乘准则的最优回归模型,并将其应用于商业银行的信贷风险评估中。实证研究表明,该方法取得的结果是可以接受的。 In view of the commercial bank＇s industrial characteristics and its problem of credit risk, it is pointed out that various kinds of data, which are depended on to forecast the performance of the commercial bank, present high nonlinearity, uncertainty and inaccuracy. These characteristics leads to inapplicability of existing evaluating models for credit risk. Thus, combined with research results of statistic learning theory, the optimal regress model based on least-absolute criteria, or LaOR model was proposed to solve the problem. Finally, an experiment was conducted to verify its validity.

关 键 词： 统计学习理论 最小一乘准则 信贷风险 商业银行

领　　域： [经济管理]