基于增量式遗传算法的粗糙集分类规则挖掘

从规则获取和优化两个方面研究了基于遗传算法(GA)的增量式粗糙集分类规则挖掘方法．通过研究决策表和决策规则系数，建立了基于粗糙集表示和度量的知识理论，将GA和粗糙集分类规则挖掘算法相结合，在保持原有知识完备的前提下，利用GA对以增量形式获得的分类规则进行优化，获取最优分类规则．试验结果表明，执行增量式GA所需时间较执行一般GA所需时间要少，可有效完成分类规则优化的任务，同时还可提高分类的精度，使分类结果具有更好的可理解性．

Rough Set Classification Rules Mining Based on Incremental Genetic Algorithm

The rough set classification rules mining based on incremental genetic algorithm (GA) is studied from two aspects: decision rules acquisition and optimization. Knowledge theory based on rough set representation and measure is constructed according to coefficients of the decision rule and decision table. To acquire optimal classification rules, the proposed method combines GA with the rough set classification rules mining algorithm. Furthermore, the rules, in incremental form, acquired by GA are optimized. Experimental results show that it performs well in the task of optimization. Comparing with the general GA it enhances the classification precision, performs task with less run time, and more understandable result can be obtained.