摘要
变精度粗集模型拓展了经典粗集理论,可使之适应不一致数据的处理。然而已有基于变精度粗集模型的知识约简算法要依据领域先验知识来确定变精度β值,影响了算法的实用性。在讨论β值对知识约简影响的基础上,提出变精度粗集模型的β值自主式获取方法,将变精度值设置在决策表相对可辨识性的阈值附近。实验结果表明该方法能较准确的反映决策表的决策分布情况。
Variable-precision rough sets model （VPRSM） is an extension of classical rough sets theory and can deal with inconsistent data. However, the parameter value of β is obtained from prior domain knowledge in existing algorithms of knowledge reduction based on VPRSM, which restricts the applications of the algorithms. The influences of variable-precision value on knowledge reduction were discussed, and an approach for self-determining β-Value in VPRSM was proposed, which sets β-value nearby the threshold value of the relative discernibility of a decision table. The experiment results show that the approach can more precisely reflect the decision distribution in the decision table.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2007年第11期2555-2558,2566,共5页
Journal of System Simulation
基金
安徽省自然科学基金(070412061)
国家自然科学基金(60575023)
博士学科点专项基金(20050359012)。
关键词
变精度粗集模型
变精度值
知识约简
决策表
variable precision rough set model
variable precision value
knowledge reduction
decision table
