基于条件信息量的动态属性约简方法

Methods of Dynamic Attribute Reduction Based on Information Quantity

针对动态建立信息系统的需要,提出了一种动态求条件信息量的方法,在一个信息系统不断变化时,该算法不像静态约简时那样需要重新计算,而是利用新增的对象对原有的信息量进行修正,利用原有信息量的结果递归计算信息系统变化后的信息量,大大节省了计算量,提高了效率。通过一个实例表明,该算法利用较小的信息表就可以计算日益庞大信息表的信息量,分析表明该算法是正确有效的。

Information quantity is the effective quantity method which delineates the classified knowledge. The computing question of the attribute reduction of information system can be turned into the calculation of the information quantity. This paper proposes a new dynamic computing information quantity algorithm. When the number of the object in the information table increases, instead of treating the changed information table as a new one and computing the information quantity again like rough set reduction algorithm does, the dynamic computing information quantity algorithm just updates the old information quantity based on the increased objects, so the computation time is greatly saved. An illustrated example shows that by the dynamic computing information quantity algorithm, the same information quantity is computed with the less objects, furthermore the results are right and effective.