粗集中规则提取的一种增量式算法

本文对数据成时间序列的动态决策表 ,用增量式算法提取决策表的规则模型。增量式算法的关键点是在分类正确率和相容度下对属性集进行优级排序 ,算法的特点是随着数据的增加逐次推导规则。随着计算轮次的增加 ,比较运算的次数按多项式增加。由于大型静态数据库可转化为动态数据库 ,所以 ,增量式算法是大型决策表规则提取的有效算法

An incremental algorithm of extracting rules in the rough set

In this paper, we extract rules of the decision table by an incremental algorithm for the dynamic decision table of the time series data. The key of the algorithm is that the attribute set is sequenced according to the classification correct ratio and the consistent degree. The distinguishing features of the algorithm is that the rules are inferred step by step with increasing data and the amounts of comparision operation increas in polynomial with increasing of the steps. Because a static large scale data base can be coverted to a dynamic data base, the incremental algorithm is a efficient algorithm of extracting rules in the large scale decision table.