王杰, 郭晨龙. 小波核极限学习机分类器[J]. 微电子学与计算机, 2013, 30(10): 73-76,80.
引用本文: 王杰, 郭晨龙. 小波核极限学习机分类器[J]. 微电子学与计算机, 2013, 30(10): 73-76,80.
WANG Jie, GUO Chen-long. Wavelet Kernel Extreme Learning Machine Classifier[J]. Microelectronics & Computer, 2013, 30(10): 73-76,80.
Citation: WANG Jie, GUO Chen-long. Wavelet Kernel Extreme Learning Machine Classifier[J]. Microelectronics & Computer, 2013, 30(10): 73-76,80.

小波核极限学习机分类器

Wavelet Kernel Extreme Learning Machine Classifier

  • 摘要: 分析了核极限学习机的原理,提出了一种小波核极限学习机,将小波函数做为极限学习机的核函数,证明了它是一种允许的极限学习机核。通过在双螺旋数据上的测试表明,小波核极限学习机在无训练数据分布的空间也具有分类能力,而高斯核极限学习机在没有训练数据分布的空间不具备分类能力。通过在不同的UCI数据集中的测试得出小波核极限学习机具有较高的分类性能。最后将小波核极限学习机应用到了人脸识别问题上,同样取得了优良的性能,说明小波核极限学习机具有一定的应用价值。

     

    Abstract: In this paper,the principle of the kernel extreme learning machine was analyzed,a kind of wavelet kernel extreme learning machine was put forward.Wavelet function was used as extreme learning machine's kernel function,and it was proved that wavelet kernel is a kind of allowed extreme learning machine kernel.Through the double helix data tests showed that wavelet kernel extreme learning machine also has the ability of classification in the absence of training data distribution,and Gauss kernel extreme learning machine don't have the ability of classification in the absence of training data distribution.Through the different UCI data set tests showed that wavelet kernel extreme learning machine has good classification performance.Finally when applied to face recognition,the wavelet kernel extreme learning machine also achieves excellent performance,showing that the wavelet kernel extreme learning machine has application value to some degree.

     

/

返回文章
返回