Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Random Fourier Features for Operator-Valued Kernels
(Submitted on 9 May 2016 (v1), revised 23 May 2016 (this version, v2), latest version 24 May 2016 (v3))
Abstract: Devoted to multi-task learning and structured output learning, operator-valued kernels provide a flexible tool to build vector-valued functions in the context of Reproducing Kernel Hilbert Spaces. To scale up these methods, we extend the celebrated Random Fourier Feature methodology to get an approximation of operator-valued kernels. We propose a general principle for Operator-valued Random Fourier Feature construction relying on a generalization of Bochner's theorem for translation-invariant operator-valued Mercer kernels. We prove the uniform convergence of the kernel approximation for bounded and unbounded operator random Fourier features using appropriate Bernstein matrix concentration inequality. An experimental proof-of-concept shows the quality of the approximation and the efficiency of the corresponding linear models on example datasets.
Submission history
From: Romain Brault Rb [view email][v1] Mon, 9 May 2016 11:36:40 GMT (383kb,D)
[v2] Mon, 23 May 2016 16:35:41 GMT (383kb,D)
[v3] Tue, 24 May 2016 12:59:58 GMT (383kb,D)
Link back to: arXiv, form interface, contact.