Kernel extrapolation

doi:10.1016/j.neucom.2005.12.113

2006-03-01

PRE

S. V. N. Vishwanathan

Borgwardt, KM

Omri Guttman

Alex Smola

DOI：10.1016/j.neucom.2005.12.113

原文链接

原文求助

摘要

En 中文

We present a framework for efficient extrapolation of reduced rank approximations, graph kernels, and locally linear embeddings (LLE) to unseen data. We also present a principled method to combine many of these kernels and then extrapolate them. Central to our method is a theorem for matrix approximation, and an extension of the representer theorem to handle multiple joint regularization constraints. Experiments in protein classification demonstrate the feasibility of our approach. (c) 2006 Elsevier B.V. All rights reserved.

Keyword:

kernel methods

regularization

graph kernels

protein classification