基于Rough集和概率统计方法的决策规则提取

Rough集是一种处理不精确的、不一致的、不完整的信息,并从中发现隐含的知识,揭示潜在规律的新型方法.Rough集理论的主要思想是在保持分类能力不变的前提下,通过属性约简,删除其中不相关或不重要的知识,从而导出问题的决策或分类规则.概率统计方法是传统数据处理方法,两者各有其优缺点.在决策规则提取过程中,将两者方法结合起来,在满足决策规则力度、肯定因子和覆盖因子的基础上对规则进行筛选,从而提高了决策的准确性和合理性.

Decision Rules Extraction Based on Rough Set and Statistics

Rough set theory is a new method of dealing with imprecise and inconsistent information,and we can discover connotative information and post potential rules with it.The primary idea in rough set theory is to educe the rules of decision or classification through reducing the attributes and deleting irrelevant or unimportant knowledge on the premise of keeping the ability of classification.Statistics is a traditional way to deal with data.Both of them have merits and shortcomings.In this paper,we bring forward a way to combine both methods in the course of decision rules extraction and the resulting-rules are sifted by support,certainty and coverage,which can improve the correctness and rationality in decision-making.