一种决策形式背景中挖掘决策规则的新方法  

A method for mining decision rules based on decision formal context

作　　者：毛华[1] 郑珍 刘晓庆[1] MAO Hua;ZHENG Zhen;LIU Xiaoqing(College of Mathematics and Information Science, Hebei University, Baoding 071002,China)

机构地区：[1]河北大学数学与信息科学学院,河北保定071002

出　　处：《河北大学学报（自然科学版）》2021年第1期1-6,共6页Journal of Hebei University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61572011);河北省自然科学基金资助项目(A2018201117)。

摘　　要：针对大规模决策形式背景中决策规则的提取问题,提出了一种将频繁闭项集与决策形式背景相结合以提取无冗余决策规则的新算法.该算法结合了深度优先搜索的思想以及概念格中概念节点之间的父子关系,改进了概念格中频繁闭项集的挖掘算法,得到的频繁决策概念格实现了对一定规模的数据库的压缩,同时得到无冗余决策规则.算法分析表明,该算法复杂度较低,更适用于大规模数据集中挖掘无冗余决策规则.Aiming at the problem of extraction of decision rules in mass decision formal context,this paper proposes a novel algorithm by combining frequent closed itemsets and decision formal context to extract no redundant decision rules.The algorithm combines the idea of depth-first searches(DFS)and the relationship of conceptual node in concept lattice and improves an existed algorithm of mining frequent closed itemsets in the concept lattice.Using this algorithm,the frequent decision concept lattice of the decision formal context is obtained so that the huge database is compressed.Meanwhile,no redundant decision rules are extracted.After analysis,this algorithm has a lower complexity and is easy to mine decision rules in decision formal context.

关 键 词：频繁闭项集 概念格 频繁决策概念格 决策规则 决策形式背景 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]