|本期目录/Table of Contents|

[1]吕林霞,赵锡英,刘光明.决策表规则提取中分辨矩阵的降级算法研究[J].工业仪表与自动化装置,2015,(02):103-107.
　L? Linxia,ZHAO Xiying,LIU Guangming.Research of algorithm for discernible matrix descending order in rule extraction of decision table[J].Industrial Instrumentation & Automation,2015,(02):103-107.
点击复制

决策表规则提取中分辨矩阵的降级算法研究(PDF)

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2015年02期
页码:
103-107
栏目:
出版日期:
2015-04-15

文章信息/Info

Title:
Research of algorithm for discernible matrix descending order in rule extraction of decision table
文章编号:
1000-0682(2015)02-0000-26
作者:
吕林霞1; 赵锡英1; 刘光明2
（1. 兰州工业学院 软件工程学院，兰州 730050；2. 甘肃省计算中心，兰州 730030）
Author(s):
L? Linxia1;  ZHAO Xiying1;  LIU Guangming2
（1. School of Software Engineering, Lanzhou Institute of Technology, Lanzhou Gansu 730050, China;2. Gansu Computing Center, Lanzhou 730030, China）
关键词:
规则提取; 分辨矩阵; 属性约简; 决策表; 粗糙集
Keywords:
rule extraction;  discernible matrix;  attribute reduction;  decision table;  rough set
分类号:
TP18
DOI:
-
文献标志码:
A
摘要:
针对大型决策表规则提取中分辨矩阵级数高，计算复杂和低效的问题，提出了一种分辨矩阵的降级算法。算法以决策等价类为对象，构建简化分辨矩阵，其中每个元素都是由两个决策等价类构成的子分辨矩阵，与传统分辨矩阵相比，简化分辨矩阵的级数低、规模小，从而简化了属性约简，效率高；提取规则时根据需要灵活设置可信度，按照可信度值提取有效规则，使算法对不一致决策表也具有较好的适应性。实例计算表明算法清晰、简捷、有效。
Abstract:
Since discernible matrix is of higher order and rule extraction computing is complicated and inefficient in large decision table, an algorithm for discernible matrix descending order was proposed. Taking decision equivalence classes as objects, the algorithm builds brief discernible matrix. Every element of the brief discernible matrix is a subordinate discernible matrix constructed by two equivalence classes. The brief discernible matrix is of lower order and smaller scale than traditional discernible matrix. The attribute reduction is simplified and efficient. When extracting rules, according to the need, the algorithm can flexibly set reliability and select effective rules by reliability value. This can make the algorithm be of better adaptability to the inconsistent decision table. Example calculation shows that the algorithm is clear, simple and effective.

参考文献/References:

[1] PAWLAK Z.Rough Set[J].International Journal of Parallel and Programming,1982,11(15):341-356.

[2] 刘文军,郭庆.基于粗糙集与模糊集的连续值域决策表的离散化算法[J].模糊系统与数学,2011,25(3):147-152.
[3] 华尔天,刘科红,肖军军,陈颖.基于粗糙集理论的设计规则提取及其应用[J].计算机工程与应用,2013,49(5): 116-119.
[4] Bahga A, Madisetti V K. Analyzing Massive Machine Maintenance Data in a Computing Cloud[J]. IEEE Transactions on Parallel and Distributed System,2012, 23(10):1831-1843.
[5] 梁德翠,胡培.面向属性归纳下的多层次决策规则获取算法[J].信息与控制,2012,41(1):69-74.
[6] 朱国龙,李龙澍.不相容决策系统中的规则提取方法研究[J].计算机工程,2013,39(2):46-49.
[7] 翟永健,张宏.优势关系多粒度粗糙模糊集及决策规则获取[J].江苏大学学报:自然科学版, 2012,33(3): 310-315.
[8] 张清华,王国胤,刘显全.基于最大粒的规则获取算法[J].模式识别与人工智能,2012,25(3):388-396.
[9] Huang C-C, Tseng B, Fan Yu-neng. Alternative rule induction methods based on incremental object using rough set theory[J].Applied Soft Computing, 2013, 13(1): 372-389.
[10] 徐怡,李龙澍,李学俊.改进的LEM2规则提取算法[J].系统工程理论与实践,2010,30(10):1841-1849.
[11] Chen Chun-Hsien, Rao Zhiming. MRM：A matrix representation and mapping approach for knowledge acquisition[J]．Knowledge Based Systems, 2008, 21(4): 284-293.
[12] 鄂旭,邵良杉,张毅智,等.一种基于粗糙集理论的规则提取方法[J].计算机科学,2011,38(1):232-239.
[13] 钱文彬,杨炳儒,徐章艳等.基于差别矩阵的不一致决策表规则获取算法[J].计算机科学,2013,40(6):215-218.

相似文献/References:

备注/Memo

备注/Memo:

收稿日期： 2014-09-26

基金项目：甘肃省自然科学研究基金计划资助项目 (1208RJZA186); 甘肃省技术研究与开发专项计划资助项目 (1205TCYA037)

作者简介：吕林霞 (1964) ，女，教授，学士，主要研究方向为数据库与优化技术 , 数据挖掘。

更新日期/Last Update: 1900-01-01