| 变量多重相关性对主成分分析的危害 |
| |
| 引用本文： | 王惠文.变量多重相关性对主成分分析的危害[J].北京航空航天大学学报,1996,22(1):65-70. |
| |
| 作者姓名： | 王惠文 |
| |
| 作者单位： | 北京航空航天大学管理学院 |
| |
| 摘 要： | 证实了主成分分析完全无法消除变量系统的多重相关性和重叠信息不良影响，指出变量多重相关性无论从方向上还是从数量上都会扭曲主成分分析的客观结论，由此得知，在进行主成分分析之前，对变量系统的选择必须是极其慎重的。
|
| 关 键 词： | 主成分分析 变量 相关性 系统分析 |
ELEMENTARY INTRODUETION TO MUTIPLE CORRELATION CAUSING HARM TO PRINCIPLE COMPONENT ANALYSIS |
| |
| Wang Huiwen.ELEMENTARY INTRODUETION TO MUTIPLE CORRELATION CAUSING HARM TO PRINCIPLE COMPONENT ANALYSIS[J].Journal of Beijing University of Aeronautics and Astronautics,1996,22(1):65-70. |
| |
| Authors: | Wang Huiwen |
| |
| Abstract: | By using Principal Component Analysis,it is unable to avoid harmful affects influenced by mutiple correlation and superimposing information within the varible system.It shows that either in principle direction or in quantity,the mutiple cornelation of the variable system will distort the results from Principal Component Analysis.Therefore,caution must be exercised in selecting varibles before using PCA. |
| |
| Keywords: | Principal Component Analysis(PCA) variation coefficient correlations systems analysis |
| 本文献已被 CNKI 维普 等数据库收录！ |
