一种改进的案例推理分类方法研究

特征属性的权重分配和案例检索策略对案例推理（Case-based reasoning，CBR）分类的准确率有显著影响. 本文提出一种结合遗传算法、内省学习和群决策思想改进的CBR分类方法. 首先，利用遗传算法得到多组属性权重，再根据内省学习原理对每组权重进行迭代调整；然后，通过案例群检索策略得到满足大多数原则的群决策分类结果；最后，以典型分类数据集的对比实验证明了本文方法能进一步提高CBR分类的准确率. 这表明内省学习可以保证权重分配的合理性，案例群检索策略能充分利用案例库的潜在信息，对提升CBR的学习能力有显著作用.

An Improved Classification Approach by Case-based Reasoning

The distribution of feature attribute weights and the strategy of case retrieval have significant impacts on the classification accuracy of case-based reasoning (CBR). An improved CBR classification approach is proposed, which is combined with genetic algorithms, introspective learning, and group decision-making theory. First, multiple attribute weights are given by a genetic algorithm. Then each group of weights is iteratively adjusted in accordance with the introspective learning principle. After that the group decision-making retrieval result which satisfies the plurality rule can be obtained according to the case group-retrieval strategy. At last, the classification comparison experiments prove that the proposed method could improve the classification accuracy of CBR. The results indicate that the introspective learning could guarantee the rationality of weight allocation, and that the case group-retrieval strategy could make full use of the potential knowledge of case base, having remarkable effects on promoting the learning ability of CBR.