stat.MLNov 10, 2026

Don't Match the Noise: Distribution Matching under Unknown Measurement Error with an Audit Sample

Authors: Arya Farahi, Ritwik Vashistha

Organizations: Department of Statistics and Data Science, University of Texas at Austin, Austin, Texas 78712, USA

Abstract

Real-world machine learning systems often operate on imperfect measurements: sensors are noisy, annotations are produced by non-experts, and scientific observations are often indirect. In these settings, classic distribution comparison tools such as maximum mean discrepancy (MMD) can be systematically biased, matching artifacts of the measurement process rather than the underlying mechanisms of interest. We study a practical regime in which noisy observations are abundant, but only a small audited subset contains ground-truth measurements. We propose measurement-error-corrected MMD (MEC-MMD), an audit-based estimator that debiases kernel evaluations without requiring a noise model. MEC-MMD is exactly unbiased for the oracle MMD under simple random auditing, admits an explicit variance decomposition, and is consistent for the population MMD, achieving the standard Op(N−1/2)O_p(N^{-1/2}) rate when the audit size scales with the dataset. We demonstrate the utility of distributional matching with MEC-MMD loss for both parameter estimation and unsupervised applications, showing that MEC-MMD consistently outperforms MMD and alternative methods. These results provide a principled foundation for robust distribution comparison, simulation-based inference, and learning in noisy real-world machine learning pipelines.

Figures & tables

Appendix figures & tables6 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Robust-by-Design Distributional Learning from Contaminated Samples

    Oct 29, 2026Nevena Gligić, Arya FarahiData ContaminationRobust Estimation

  2. Optimal Guarantees for Auditing Rényi Differentially Private Machine Learning

    May 21, 2026Benjamin D. Kim, Lav R. Varshney, Daniel AlabiPrivacy AuditingBlack-Box Model Auditing

  3. Diffusion-based Denoising Beats Vanilla Score Matching in Parameter Estimation: A Theoretical Explanation

    May 21, 2026Benedikt Lütke Schwienhorst, Nadja Klein, Johannes LedererScore MatchingParameter Estimation