邻域系统的不确定性度量方法

• 作 者: 唐朝辉 陈玉明
• 机 构: 厦门理工学院计算机信息与工程学院,
• 出 处: 控制与决策
• 发表日期: 2014年04期
• 关键词: 邻域系 不确定性 度量 近似精度 不确定性度量 邻域系统 粗糙集 知识粒度 粗糙集模型 等价关系

针对离散型数据系统的不确定性度量方法难以有效解决邻域系统不确定性度量的问题,引入邻域粗糙集模型,提出邻域精确度、邻域知识粒度和基于邻域知识粒度的近似精度等邻域系统不确定性度量方法,进一步从理论上证明其有效性.实验结果表明,基于邻域知识粒度的近似精度具有更严格的单调性,优于邻域近似精度的邻域系统对不确定性度量的效果.

The uncertainty measurement method for discrete data system is difficult to effectively solve the problem of uncertainty measurement of neighborhood system. By introducing the neighborhood rough set model, the neighborhood accuracy, the neighborhood knowledge granularity and the approximate accuracy based on the neighborhood knowledge granularity are proposed And other neighborhood system uncertainty measurement methods to further prove its validity theoretically.The experimental results show that the approximate accuracy based on the neighborhood knowledge granularity has a more strict monotonicity and the neighborhood system that is better than the neighborhood approximate accuracy does not The effect of certainty measure.