# General regularization in covariate shift adaptation

arXiv: https://arxiv.org/abs/2307.11503 · HTML: https://mlsanity.com/paper/2307.11503/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 40 references that are in the map.
- First posted on arXiv 21 Jul 2023.

## Builds on (most cited first, 7 of 7)

- [Learning in Implicit Generative Models](https://mlsanity.com/paper/1610.03483/) (2016)
- [Less is More: Nystr"om Computational Regularization](https://mlsanity.com/paper/1507.04717/) (2015)
- [Addressing Parameter Choice Issues in Unsupervised Domain Adaptation by Aggregation](https://mlsanity.com/paper/2305.01281/) (2023)
- [Kernel Conditional Density Operators](https://mlsanity.com/paper/1905.11255/) (2019)
- [Relating Leverage Scores and Density using Regularized Christoffel Functions](https://mlsanity.com/paper/1805.07943/) (2018)
- [Inverse Density as an Inverse Problem: The Fredholm Equation Approach](https://mlsanity.com/paper/1304.5575/) (2013)
- [Analysis of regularized Nystr"om subsampling for regression functions of low smoothness](https://mlsanity.com/paper/1806.00826/) (2018)

## Built on by (most cited first, 2 of 2)

- [Spectral Algorithms under Covariate Shift](https://mlsanity.com/paper/2504.12625/) (2025)
- [Rethinking Data Augmentation under Covariate Shift: Invariant-Guided Diffusion and Prototype Reweighting](https://mlsanity.com/paper/2610.00873/) (2026)

## Abstract (arXiv)

Sample reweighting is one of the most widely used methods for correcting the error of least squares learning algorithms in reproducing kernel Hilbert spaces (RKHS), that is caused by future data distributions that are different from the training data distribution. In practical situations, the sample weights are determined by values of the estimated Radon-Nikod'ym derivative, of the future data distribution w.r.t.~the training data distribution. In this work, we review known error bounds for reweighted kernel regression in RKHS and obtain, by combination, novel results. We show under weak smoothness conditions, that the amount of samples, needed to achieve the same order of accuracy as in the standard supervised learning without differences in data distributions, is smaller than proven by state-of-the-art analyses.