Using the Equivalent Kernel to Understand Gaussian
Peter Sollich and Dr Christopher Williams
In: Neural Information Processing Systems 2004 (NIPS 2004), 14-16 Dec 2004, Vancouver, Canada.
The equivalent kernel (Silverman, 1984) is a way of understanding
how Gaussian process regression works for large sample sizes based on
a continuum limit. In this
paper we show (1) how to approximate the equivalent kernel of the
widely-used squared exponential (or Gaussian) kernel and
related kernels, and (2) how analysis using the equivalent kernel
helps to understand the learning curves for Gaussian processes.