论文信息 - Using the Equivalent Kernel to Understand Gaussian Process Regression

Using the Equivalent Kernel to Understand Gaussian Process Regression

The equivalent kernel [1] 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.

Christopher K. I. Williams | Peter Sollich | Peter Sollich

[1] Athanasios Papoulis,et al. Probability, Random Variables and Stochastic Processes , 1965 .

[2] Bruno O. Shubert,et al. Random variables and stochastic processes , 1979 .

[3] M. V. Rossum,et al. In Neural Computation , 2022 .

[4] Tomaso A. Poggio,et al. Regularization Theory and Neural Networks Architectures , 1995, Neural Computation.

[5] Manfred Opper,et al. General Bounds on Bayes Errors for Regression with Gaussian Processes , 1998, NIPS.

[6] Alexander J. Smola,et al. Learning with kernels , 1998 .

[7] Peter Sollich,et al. Learning Curves for Gaussian Process Regression: Approximations and Bounds , 2001, Neural Computation.

[8] D. Hinkley. Annals of Statistics , 2006 .