| 小样本问题的算法比较 |
| |
| 引用本文： | 张先荣,范丽亚.小样本问题的算法比较[J].聊城大学学报(自然科学版),2011(2):23-27. |
| |
| 作者姓名： | 张先荣 范丽亚 |
| |
| 作者单位： | 聊城大学数学科学学院; |
| |
| 基金项目： | 国家自然科学基金(10871226); 山东省自然科学基金资助项目(ZR2009AL006) |
| |
| 摘 要： | 将不相关线性判别分析(ULDA)和零空间线性判别分析(NLDA)两种思想结合起来,提出了处理小样本问题的六种算法,并通过实验说明了这六种算法的分类有效性.
|
| 关 键 词： | 不相关线性判别分析 零空间线性判别分析 特征值分解 错分率 |
Comparison of Algorithms for the Small Sample Size Problems |
| |
| ZHANG Xian-rong FAN Li-ya.Comparison of Algorithms for the Small Sample Size Problems[J].Journal of Liaocheng University:Natural Science Edition,2011(2):23-27. |
| |
| Authors: | ZHANG Xian-rong FAN Li-ya |
| |
| Institution: | ZHANG Xian-rong FAN Li-ya (School of Mathematical Sciences,Liaocheng University,Liaocheng 252059,China) |
| |
| Abstract: | By combining the ideas of NLDA and ULDA,in this paper,six new algorithms for the small sample size problems are presented.Experiments are provided for illustrating the classification effectiveness of the six algorithms presented in the paper with Yale B and AR face data sets. |
| |
| Keywords: | uncorrelated discriminant analysis null-space discriminant analysis eigen-value decomposition misclassification rate |
| 本文献已被 CNKI 等数据库收录！ |
