English

Validation design I: construction of validation designs via kernel herding

Methodology 2021-12-13 v1

Abstract

We construct validation designs ZmZ_m aimed at estimating the integrated squared prediction error of a given design XnX_n. Our approach is based on the minimization of a maximum mean discrepancy for a particular kernel, conditional on XnX_n, so that sequences of nested validation designs can be constructed incrementally by kernel herding. Numerical experiments show that key features for a good validation design are its space-filling properties, in order to fill the holes left by XnX_n and properly explore the whole design space, and the suitable weighting of its points, since evaluations far from XnX_n tend to overestimate the global error. A dedicated weighting method, based on a particular kernel, is proposed. Numerical simulations with random functions show the superiority the method over more traditional validation based on random designs, low-discrepancy sequences, or leave-one-out cross validation.

Keywords

Cite

@article{arxiv.2112.05583,
  title  = {Validation design I: construction of validation designs via kernel herding},
  author = {Luc Pronzato and Maria-João Rendas},
  journal= {arXiv preprint arXiv:2112.05583},
  year   = {2021}
}
R2 v1 2026-06-24T08:12:22.432Z