English

Bayesian leave-one-out cross-validation for large data

Machine Learning 2020-08-12 v1 Machine Learning

Abstract

Model inference, such as model comparison, model checking, and model selection, is an important part of model development. Leave-one-out cross-validation (LOO) is a general approach for assessing the generalizability of a model, but unfortunately, LOO does not scale well to large datasets. We propose a combination of using approximate inference techniques and probability-proportional-to-size-sampling (PPS) for fast LOO model evaluation for large datasets. We provide both theoretical and empirical results showing good properties for large data.

Keywords

Cite

@article{arxiv.1904.10679,
  title  = {Bayesian leave-one-out cross-validation for large data},
  author = {Måns Magnusson and Michael Riis Andersen and Johan Jonasson and Aki Vehtari},
  journal= {arXiv preprint arXiv:1904.10679},
  year   = {2020}
}

Comments

Accepted to ICML 2019. This version is the submitted paper

R2 v1 2026-06-23T08:48:03.350Z