| 基于容错改进的邻域粗糙集属性约简算法 |
| |
| 引用本文： | 彭潇然,刘遵仁,纪俊.基于容错改进的邻域粗糙集属性约简算法[J].计算机应用研究,2018,35(8). |
| |
| 作者姓名： | 彭潇然 刘遵仁 纪俊 |
| |
| 作者单位： | 青岛大学数据科学与软件工程学院,青岛大学计算机科学技术学院,青岛大学计算机科学技术学院 |
| |
| 基金项目： | 国家自然科学基金资助项目（61503208） |
| |
| 摘 要： | 作为Pawlak粗糙集的扩展，邻域粗糙集能有效地处理数值型的数据。但是，因为沿用了Pawlak粗糙集在构造上下近似集时的包含关系，邻域粗糙集对噪声数据的容错性很差。针对这个问题，本文通过引入贝叶斯最小风险决策规则，提出了一种基于容错改进的邻域粗糙集属性算法。通过和现有的算法进行比较，实验结果表明，在数据预处理阶段用该算法能得到更好的属性约简。
|
| 关 键 词： | 粗糙集 邻域粗糙集 决策粗糙集 属性约简 容错性 |
| 收稿时间： | 2017/4/12 0:00:00 |
| 修稿时间： | 2018/7/2 0:00:00 |
Attribute Reduction Algorithm Based on Fault-Tolerance Improvement of Neighborhood Rough Set |
| |
| peng xiao ran,liu zun ren and ji jun.Attribute Reduction Algorithm Based on Fault-Tolerance Improvement of Neighborhood Rough Set[J].Application Research of Computers,2018,35(8). |
| |
| Authors: | peng xiao ran liu zun ren and ji jun |
| |
| Affiliation: | College of Data Science and Software Engineering, Qingdao University,, |
| |
| Abstract: | As the extension of Pawlak rough set, neighborhood rough set can effectively deal with numerical data. However, its fault tolerance is very poor to noise data, because it follows the inclusion relation which is used for constructing the upper and lower approximations in Pawlak rough set. In order to solve this problem, this paper presents a new algorithm based on fault-tolerance improvement of neighborhood rough set by introducing the Bayes decision with minimum risk. Compared with the existing algorithm, the experimental results show that the attribute reduction obtained by this proposed algorithm is better in the data pre-processing. |
| |
| Keywords: | |
|
| 点击此处可从《计算机应用研究》浏览原始摘要信息 |
|
点击此处可从《计算机应用研究》下载全文 |
|
