针对相容决策表,另辟一种数据约简思路,与传统数据约简方法不同的是,本文方法考虑属性值约简与属性约简之间的关联,该方法首先进行属性值约简,然后再进行属性约简,本文并给予了充分的证明,而普通的方法并没有考虑它们之间的联系,并有许多冗余的比较操作。
As for the consistent decision table, this paper proposes a new method of data reduction. Compared with traditional da- ta reduction method, this paper considers the relation between attribute reduction and attribute value reduction, that is, the meth- od firstly gives attribute value reduction, and then conducts attribute reduction, last this paper proves it. However, lots of general method of data reduction, they don't consider the relation and much redundant operations.