论文部分内容阅读
置信优势关系粗糙集是处理不完备有序信息的重要模型,上、下近似集的计算是核心内容之一.在实际应用中,属性集通常会发生变化.根据属性集的增加或减少,首先讨论置信优势类及劣势类变化情况,随之给出上、下近似集增量式的变化规律,提出相应的近似集动态更新方法.通过Matlab在UCI数据集上的实验结果表明,与非增量式方法相比,所提出的置信优势关系粗糙集下的上、下近似集的增量式更新方法可行、高效.
Confidence Dominance Rough set is an important model to deal with incomplete ordered information, the calculation of the upper and lower approximation sets is one of the core content.In the practical application, the attribute set usually changes.According to the increase or decrease of attribute set, Discuss the changes of confidence superiority class and inferiority class, then give the variation rule of upper and lower approximation sets increment, and put forward corresponding dynamic update method of approximation sets.Experimental results on UCI datasets by Matlab show that, Compared with the quantitative method, the incremental updating method of the upper and lower approximation sets under the proposed confidence-dominant relational rough set is feasible and efficient.