| Reducing Inconsistent Rules Based on Irregular Decision Table |
| |
| 作者姓名： | 兰轶东 张霖 刘连臣 |
| |
| 作者单位： | [1]DepartmentofMechanicsandEngineeringScience,PekingUniversity,Beijing100871,China [2]CIMSEngineeringResearchCenter,TsinghuaUniversity,Beijing100084,China |
| |
| 基金项目： | the Basic Research Foundation of Tsinghua University (No. JC2001029) and the National High-Tech Research and Development Program of China (No. 863-511-930-004) |
| |
| 摘 要： | In this paper, we study the problem of rule extraction from data sets using the rough set method. For inconsistent rules due to improper selection of split-points during discretization, and/or to lack of information, we propose two methods to remove their inconsistency based on irregular decision tables. By using these methods, inconsistent rules are eliminated as far as possible, without affecting the remaining consistent rules. Experimental test indicates that use of the new method leads to an improvement in the mean accuracy of the extracted rules.
|
| 关 键 词： | 不规则决策表 粗糙集 分流点 数据提取 |
Reducing Inconsistent Rules Based on Irregular Decision Table |
| |
| LAN Yidong,ZHANG Lin,LIU Lianchen ?.Reducing Inconsistent Rules Based on Irregular Decision Table[J].Tsinghua Science and Technology,2004,9(1):45-50. |
| |
| Authors: | LAN Yidong ZHANG Lin LIU Lianchen ? |
| |
| Institution: | LAN Yidong,ZHANG Lin,LIU Lianchen )?** Department of Mechanics and Engineering Science,Peking University,Beijing 100871,China, ?CIMS Engineering Research Center,Tsinghua University,Beijing 100084,China |
| |
| Abstract: | In this paper, we study the problem of rule extraction from data sets using the rough set method. For inconsistent rules due to improper selection of split-points during discretization, and/or to lack of information, we propose two methods to remove their inconsistency based on irregular decision tables. By using these methods, inconsistent rules are eliminated as far as possible, without affecting the remaining consistent rules. Experimental test indicates that use of the new method leads to an improvement in the mean accuracy of the extracted rules. |
| |
| Keywords: | rough set inconsistent rule irregular decision table discretization split-point classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
