English

Convex regularization in statistical inverse learning problems

Machine Learning 2021-11-02 v3 Machine Learning Statistics Theory Statistics Theory

Abstract

We consider a statistical inverse learning problem, where the task is to estimate a function ff based on noisy point evaluations of AfAf, where AA is a linear operator. The function AfAf is evaluated at i.i.d. random design points unu_n, n=1,...,Nn=1,...,N generated by an unknown general probability distribution. We consider Tikhonov regularization with general convex and pp-homogeneous penalty functionals and derive concentration rates of the regularized solution to the ground truth measured in the symmetric Bregman distance induced by the penalty functional. We derive concrete rates for Besov norm penalties and numerically demonstrate the correspondence with the observed rates in the context of X-ray tomography.

Keywords

Cite

@article{arxiv.2102.09526,
  title  = {Convex regularization in statistical inverse learning problems},
  author = {Tatiana A. Bubba and Martin Burger and Tapio Helin and Luca Ratti},
  journal= {arXiv preprint arXiv:2102.09526},
  year   = {2021}
}

Comments

35 pages, 4 figures

R2 v1 2026-06-23T23:18:00.294Z