| 两种显著性检验方法在海洋大数据序列主成分分析中的应用 |
| |
| 引用本文： | 方文东.两种显著性检验方法在海洋大数据序列主成分分析中的应用[J].热带海洋学报,1998(4). |
| |
| 作者姓名： | 方文东 |
| |
| 作者单位： | 中国科学院南海海洋研究所!广州，510301 |
| |
| 基金项目： | 广东省自然科学基金!950704，中国科学院南海海洋研究所所长基金 |
| |
| 摘 要： | 介绍海洋大数据序列主成分分析的两种显著性检验方法，采用该两种方法可以辨别具有真实意义的主成分模与代表随机噪声信号的模态。此二法均能很好地应用于大数据序列主成分模的显著性检验。在进行海洋大数据序列时-空变化主成分分析中，为尽可能减少或避免数据中随机噪声的影响，建议对主成分模进行显著性检验。
|
| 关 键 词： | 海洋数据 主成分分析 显著性检验 |
APPLICATION OF TWO SIGNIFICANCE TESTS TO PRINCIPAL COMPONENT ANALYSIS OF LARGE MARINE DATA SETS |
| |
| Fang Wendong.APPLICATION OF TWO SIGNIFICANCE TESTS TO PRINCIPAL COMPONENT ANALYSIS OF LARGE MARINE DATA SETS[J].Journal of Tropical Oceanography,1998(4). |
| |
| Authors: | Fang Wendong |
| |
| Abstract: | Two methods of significance tests for principal component (PC) analysis on large marine data sets are nitroduced. Meaningful PC modes can be distinguished from random noise PC modes by using the two methods. The two methods are good to be used for the eigenvalue significance tests of large data sets. Significance tests on PC modes are suggested to be employed in the PC analysis of bine-space variability of large marine data sets so as to reduce or avoid the effect of random noise process. |
| |
| Keywords: | marine data principal component analysis significance test |
| 本文献已被 CNKI 等数据库收录！ |
