New Probabilistic Bounds on Eigenvalues and Eigenvectors of Random Kernel Matrices

Details New Probabilistic Bounds on Eigenvalues and Eigenvectors of Random Kernel Matrices New Probabilistic Bounds on Eigenvalues and Eigenvectors of Random Kernel Matrices

2012-02-14

arxiv.org/abs/1202.3761v1

Computer Science

Learning

Scientific paper

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.

Affiliated with

Also associated with

No associations

Rating

New Probabilistic Bounds on Eigenvalues and Eigenvectors of Random Kernel Matrices does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with New Probabilistic Bounds on Eigenvalues and Eigenvectors of Random Kernel Matrices, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and New Probabilistic Bounds on Eigenvalues and Eigenvectors of Random Kernel Matrices will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-90672