| 基于可辨识矩阵的启发式属性约简方法及其应用 |
| |
| 引用本文： | 芦晓红,陈世权,吴今培.基于可辨识矩阵的启发式属性约简方法及其应用[J].计算机工程,2003,29(1):56-59. |
| |
| 作者姓名： | 芦晓红 陈世权 吴今培 |
| |
| 作者单位： | 五邑大学智能技术与系统研究所,江门,529020 |
| |
| 基金项目： | 国家自然科学基金资助项目（70071032），广东省自然科学基金资助项目（000874） |
| |
| 摘 要： | 在基于可辨识矩阵的属性约简算法的基础上，提出了基于可辨识矩阵的计算属性重要性的方法，并以此作为启发式知识来约简决策表中的冗余属性。这种方法直接源于评审数据，思路清晰，拟合结果表明本约简算法合理、可靠。
|
| 关 键 词： | 粗糙集 属性约简 可辨识矩阵 属性重要性 |
| 文章编号： | 1000-3428(2003)01-0056-04 |
| 修稿时间： | 2002年1月22日 |
Heuristic Attribute Reduction Based on Discernibility Matrix and Its Application |
| |
| LU Xiaohong,CHEN Shiquan,WU Jinpei.Heuristic Attribute Reduction Based on Discernibility Matrix and Its Application[J].Computer Engineering,2003,29(1):56-59. |
| |
| Authors: | LU Xiaohong CHEN Shiquan WU Jinpei |
| |
| Abstract: | For the purpose of making up the faultiness of attribute reduction algorithm based on discernibility matrix, the method of attribute importance based on discernibility matrix is put forward in this paper and is used as heuristic knowledge in deciding the sequence of attribute reduction. Compared with general methods, the result attained by this method rooted in evaluation data directly, so it has more reliability and strong persuasion. |
| |
| Keywords: | Rough sets Attribute reduction Discernibility matrix Attribute importance |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
