| 稳定的特征选择研究 |
| |
| 引用本文： | 李云. 稳定的特征选择研究[J]. 微型机与应用, 2012, 31(15): 1-2,5 |
| |
| 作者姓名： | 李云 |
| |
| 作者单位： | 南京邮电大学计算机学院,江苏南京,210003 |
| |
| 基金项目： | 基金项目:国家自然科学基金(61073114):南京邮电大学攀登计划(NY210010)资助课题 |
| |
| 摘 要： | 特征选择是机器学习和数据挖掘领域的关键问题之一,而特征选择的稳定性也是目前的一个研究热点。主要对特征选择的稳定性因素和稳定性度量进行分析,并详细介绍了目前比较经典的两种提高特征选择稳定性的方法。
|
| 关 键 词： | 特征选择 稳定性 集成 样本加权 |
Research on stable feature selection |
| |
| Li Yun. Research on stable feature selection[J]. Microcomputer & its Applications, 2012, 31(15): 1-2,5 |
| |
| Authors: | Li Yun |
| |
| Affiliation: | Li Yun (College of Computer,Nanjing University of Posts and Telecommunications,Nanjing 210003,China) |
| |
| Abstract: | Feature selection is one of the key problems in machine learning and data mining, and the stability of feature selection is one of the current hot points. In the paper, the factors to the stability of feature selection and the measure of stability is introduced, at the same time, two classic methods to improve the stability is presented. |
| |
| Keywords: | feature selection stability ensemble instance-weighting |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
