基于TL_1罚的稀疏主成分分析 |
| |
引用本文: | 任佳佳,李海洋,张倩.基于TL_1罚的稀疏主成分分析[J].纺织高校基础科学学报,2016(4):478-483. |
| |
作者姓名: | 任佳佳 李海洋 张倩 |
| |
作者单位: | 西安工程大学理学院,陕西西安,710048 |
| |
基金项目: | 国家自然科学基金资助项目(11271297),陕西省自然科学基金资助项目(2015JM1012) |
| |
摘 要: | 在稀疏主成分分析的基础上,用TL_1罚替代稀疏主成分分析中的L_1罚,提出TL_1罚稀疏主成分分析,并给出TL_1罚稀疏主成分分析的阈值迭代算法.最后,以蔬菜选择为例进行数值模拟,结果表明TL_1罚稀疏主成分分析效果更优.
|
关 键 词: | 稀疏主成分分析 TL1罚 阈值迭代算法 |
Sparse principal component analysis based on transformed L1 penalty |
| |
Abstract: | Transformed L1 penalty sparse principal component analysis was proposed by replacing L1 penalty with transformed L1 penalty based on sparse principal component analysis,and a thresholding iteration algorithm was given for transformed L1 penalty sparse principal component analysis.Finally,taking vegetable selection as an example,the numerical simulation results show that transformed L1 penalty sparse principal component analysis performs better. |
| |
Keywords: | sparse principal component analysis transformed L1 penalty iterative thresholding algorithm |
本文献已被 CNKI 万方数据 等数据库收录! |
|