一种基于知识距离的决策规则的评测

粒度计算是信息处理的一种新的概念和计算范式,覆盖了所有有关粒度的理论、方法、技术和工具的研究。主要用于处理模糊的、不完整的、不确定的和海量的信息。粗糙集模型作为粒度计算的三大研究模型之一,得到了广泛的应用。针对粗糙集中决策规则性能的评测问题,本文基于粒度计算的思想,提出一种粒化算法,通过知识距离来粗化条件属性集,实现对决策规则的评测。实例证明该方法评测效果较为理想。

Evaluation on a Decision Rule Based on Knowledge Distance

Grain calculation as a new concept aria caLcu,au,,e on grain theories, technologies, techniques and tools. It is mainly used to process the fuzzy, incomplete, uncertain and mass information. The model of roughness set as one of the main research models in grain calculation is widely used. According to the evaluation problems in decision rule performance in roughness set, a granulated algorithm was put forward to evaluate decision rules by coarsening condition property set through knowledge distance. Samples proved that this method has more ideal evaluation effects.