基于流图的新型规则推理网络研究及应用

Non-linear Correlation Discovery-based Algorithm in Data Mining

本文提出了一种基于流图(flow graph)的新型数据挖掘技术,用于数据中的规则提取以及推理,并且首次将该技术应用到商业银行信用风险决策中,通过数据实验证实了它的有效性。

In this paper, we presente a novel datamining method of flow graph-based rules referring network in the field of credit risk evaluation and the optimal portfolio, and study the reference character of the flow graph, then put out its application into the real data compared with recently neural network, getting outstanding achiecement.