English

The linear conditional expectation in Hilbert space

Statistics Theory 2021-08-26 v2 Functional Analysis Machine Learning Statistics Theory

Abstract

The linear conditional expectation (LCE) provides a best linear (or rather, affine) estimate of the conditional expectation and hence plays an important r\^ole in approximate Bayesian inference, especially the Bayes linear approach. This article establishes the analytical properties of the LCE in an infinite-dimensional Hilbert space context. In addition, working in the space of affine Hilbert--Schmidt operators, we establish a regularisation procedure for this LCE. As an important application, we obtain a simple alternative derivation and intuitive justification of the conditional mean embedding formula, a concept widely used in machine learning to perform the conditioning of random variables by embedding them into reproducing kernel Hilbert spaces.

Keywords

Cite

@article{arxiv.2008.12070,
  title  = {The linear conditional expectation in Hilbert space},
  author = {Ilja Klebanov and Björn Sprungk and T. J. Sullivan},
  journal= {arXiv preprint arXiv:2008.12070},
  year   = {2021}
}

Comments

32 pages

R2 v1 2026-06-23T18:08:22.993Z