| 一个基于诊断的稳健主成分分析方法 |
| |
| 引用本文： | 陈文灿,崔卉.一个基于诊断的稳健主成分分析方法[J].分析科学学报,1996,12(4):283-287. |
| |
| 作者姓名： | 陈文灿 崔卉 |
| |
| 作者单位： | 湖南大学化学计量学与化学传感器技术研究所 |
| |
| 基金项目： | 国家教委霍英东基金,国家自然科学基金 |
| |
| 摘 要： | 经典的主成分分析方法易受异常点影响。本文根据该方法的特点，提出一新的诊断方法，将多变量数据中异常剔除后再进行主成分分析，构成有效的稳健主成分分析法。用此法处理二组实际数据，结果令人满意。
|
| 关 键 词： | 稳健主成分分析 异常点 诊断 分析化学 |
A New Robust PCA Based on Diagnosis |
| |
| Chen Wencan,Cui Hui,Liang Yizeng.A New Robust PCA Based on Diagnosis[J].Journal of Analytical Science,1996,12(4):283-287. |
| |
| Authors: | Chen Wencan Cui Hui Liang Yizeng |
| |
| Abstract: | Classic principal component analysis is vulnerable to outliers. A new robust principal component analysis method based on diagnosis is presented in this paper. First, a powerful diagnostic method developed by the authors is applied to analytical data. After rejecting outliers, classic PCA is used. The method is easy to comprehend and has light calculations. |
| |
| Keywords: | Robust principal component analysis Outliers Diagnosis Chemometrics |
| 本文献已被 CNKI 维普 等数据库收录！ |
