arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1202.3761cs.LGstat.ML
收藏

随机核矩阵特征值和特征向量的新概率界

New Probabilistic Bounds on Eigenvalues and Eigenvectors of Random Kernel Matrices

Aalto University, School of Science(阿尔托大学理学院);Waseda University(早稻田大学)

Nima Reyhani, Hideitsu Hino, Ricardo Vigario

摘要:

核方法是解决不同机器学习问题的成功方法。这种成功主要源于使用特征映射和内核矩阵。一些方法依赖于核矩阵的特征值/特征向量,而对于其他方法,谱信息可以用于估计超额风险。一个重要的问题仍然是样本特征值/特征向量与总体值有多接近。本文改进了以往关于一般核矩阵特征值集中界的结果。对于距离和内积核函数,例如径向基函数,我们提供了新的浓度界,其特征在于样本协方差矩阵的特征值。同时,更清晰的边界的障碍是占和部分解决。作为一个案例研究,我们推导出样本核目标对准的浓度不等式。

英文摘要:

Kernel methods are successful approaches for different machine learning problems. This success is mainly rooted in using feature maps and kernel matrices. Some methods rely on the eigenvalues/eigenvectors of the kernel matrix, while for other methods the spectral information can be used to estimate the excess risk. An important question remains on how close the sample eigenvalues/eigenvectors are to the population values. In this paper, we improve earlier results on concentration bounds for eigenvalues of general kernel matrices. For distance and inner product kernel functions, e.g. radial basis functions, we provide new concentration bounds, which are characterized by the eigenvalues of the sample covariance matrix. Meanwhile, the obstacles for sharper bounds are accounted for and partially addressed. As a case study, we derive a concentration inequality for sample kernel target-alignment.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑