基于粗糙集的决策规则提取

Rough set-based decision rules extraction

讨论了粗糙集(rough sets)在数据分析方面的应用,基于粗糙集的数据分析来源于对数据表(即信息系统)的分析。通常我们将信息系统中的属性划分为条件属性和决策属性,这种信息系统又称为决策表。对每一个决策表,都有一系列的决策规则与之相对应,这些相关的决策规则被称为决策算法。将这些决策算法与贝叶斯(Bayes’)定理相比较,从而得出基于粗糙集的数据分析方法优于贝叶斯定理。

Rough set-based data analysis starts from a data table, called an information system. The attribute of information system is usually divided into two parts, condition attributes and decision. Such information system is called decision table. In every decision table a set of decision rules, called a decision algorithm, can be associated. It is shown that every decision algorithm reveals some well-known probabilistic properties, and we compare it with Bayes’theorem.