论文部分内容阅读
以同时具有丢失型和遗漏型未知属性值的广义不完备序值信息系统为研究对象,提出确定特征优势关系的概念,并基于此关系讨论了该系统的知识约简问题.对比分析证明,确定特征优势关系在一定程度上克服了特征优势关系的局限性.在广义不完备序值决策系统中引入近似分布约简,提出新的近似分布约简概念,结果表明基于确定特征优势关系粗糙集模型能有效地处理具有两种未知属性值的广义不完备序值系统,为从复杂的不完备系统中知识获取提供了新的理论基础与技术手段.
Taking the generalized incomplete sequence valued information system with both missing and missing unknown attribute values as the research object, this paper proposes the concept of determining the dominant relation of features and discusses the knowledge reduction problem of the system based on this relationship. To some extent, the feature predominance overcame the limitations of the feature predominance relation.The approximate distribution reduction was introduced into the generalized incomplete sequence valued decision-making system, and a new conception of approximate distribution reduction was proposed.The results show that based on the feature-based rough set model It can deal with generalized incomplete sequence valued system with two unknown attribute values effectively and provide a new theoretical basis and technical means for obtaining knowledge from complex incomplete system.