|
|||||
|
|
| 主成分分析法中数据处理方法的改进 | |
| 引用本文： | 纪荣芳. 主成分分析法中数据处理方法的改进[J]. 山东科技大学学报(自然科学版), 2007, 26(5): 95-98 |
| 作者姓名： | 纪荣芳 |
| 作者单位： | 泰山学院,数学与系统科学系,山东,泰安,271021 |
| 摘 要： | 由于原有的主成分分析法受到评价指标量纲和数量级的影响，因而主成分也会因评价指标量纲和数量级的改变而不同，并且使得部分信息丢失。用均值化方法借助DPS数据处理软件，对原始数据进行无量纲化处理，不仅有效消除了量纲与数量级的影响，而且使得处理后的数据包含了原始数据的全部信息。 |
| 关 键 词： | 主成分综合评价法 协方差矩阵 均值化 贡献率 |
| 文章编号： | 1672-3767（2007）05-0095-04 |
| 收稿时间： | 2006-10-17 |
| 修稿时间： | 2006-10-17 |
Improvement of Data Processing Method in Principal Component Analysis |
|
| JI Rong-fang. Improvement of Data Processing Method in Principal Component Analysis[J]. Journal of Shandong Univ of Sci and Technol: Nat Sci, 2007, 26(5): 95-98 | |
| Authors: | JI Rong-fang |
| Abstract: | Due to the quantities and dimensions of the evaluation indices affect the original principal component analy- sis method, so the principal component will be different with the change of quantities and dimensions of the evalua- tion indices, and some information will be lost. With the help of the data processing software DPS, we make the non-dimensional processing of the raw data by means of the equalization method. As a result, the improved principal component analysis method not only eliminates the effect of quantities and dimensions, but also includes all the in formation of raw data. |
| Keywords: | principal component synthetic evaluation method covariance matrix equalization contribution rate |
| 本文献已被 维普 万方数据 等数据库收录！ | |