摘要
核Foley-Sammon鉴别分析由于可以抽取得到原始样本的非线性正交特征,因此被广泛应用于模式识别的研究领域。但是该算法在具体求解每一个特征矢量过程中均需求解相应的广义特征方程,因此非常耗时。为了克服这一困难,提出了一种新的快速近似算法即核Foley-Sammon鉴别分析,有效地避免了多次求解广义特征方程。在ORL人脸数据库上的实验结果表明,该算法不仅在识别性能上优于核线性鉴别分析,而且在特征抽取速度上优于传统的核Foley-Sammon鉴别分析。
Kernel Foley-Sammon Discriminant Analysis (KFSDA) is extensively applied on the field of pattern recognition with the capacity of extracting nonlinear orthonormal features from original samples. But each eigenvector should be achieved by calculating the corresponding generalized eigenfunction with this algorithm, and it is very time-consuming. In order to overcome this problem, a fast algorithm of KFSDA named Fast Kernel Foley-Sammon Discriminant Analysis (FKFSDA) was proposed. Experimental results on the ORL face database demonstrate the proposed algorithm not only outperforms Kernel Linear Discriminant Analysis (KLDA) in recognition rates, but also outperforms conventional Foley-Sammon Discriminant Analysis in feature extraction speed.
出处
《计算机科学》
CSCD
北大核心
2009年第6期273-275,285,共4页
Computer Science
基金
国家自然科学基金(60572034)
江苏省自然科学基金(BK2004058)
江苏科技大学青年教师科研立项资助