一种新的分类规则提取算法  

Novel Classification Rule Acquisition Algorithm

作　　者：赵旭俊[1] 张继福[1] 马洋[1] 蔡江辉[1] 

机构地区：[1]太原科技大学计算机科学与技术学院,太原030024

出　　处：《小型微型计算机系统》2012年第5期1126-1130,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(60773014;41140027)资助;国家自然基金主任基金项目(41140027)资助;山西省青年科技基金项目(2008021028)资助

摘　　要：约束概念格是概念格的特化结构,构造时具有较低的时空复杂度,能从中快速提取比较丰富的信息和知识.为了提取分类规则,在充分分析约束概念格结点外延与数据集等价划分之间关系的前提下,引入了分类支持度和记录支持度的概念,提出了一种面向约束概念格的分类规则提取算法(Classification Rule Acquisition Algorithm based on Constrained Concept Lattice,CRACCL),并采用UCI数据集作为实验集,验证了本算法能够提取更加实用和准确的分类规则.Constrained concept lattice,with the characteristics of higher constructing efficiency,practicability and pertinence,is a new concept lattice structure.Abundant information and knowledge can be quickly extracted from concept lattice.For classification rule acquisition,a classification rule acquisition algorithm CRACCL based on the constrained concept lattice is presented by using the concept of classification support and record support according to the relationship between node′s extent of constrained concept lattice and equivalence partition of data set.The experiment results validate the higher classification efficiency and correctness of the algorithm by taking the UCI data sets as the formal contexts.

关 键 词：约束概念格 分类规则 分类支持度 记录支持度 相容结点 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]