Disagreement-Regularized Importance Sampling for Adversarial Label Corruption

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Horváth, Csongor, Sintorn, Ida-Maria, Singh, Prashant
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918489366724608
author Horváth, Csongor
Sintorn, Ida-Maria
Singh, Prashant
author_facet Horváth, Csongor
Sintorn, Ida-Maria
Singh, Prashant
contents Standard Importance Sampling (IS) collapses under label corruption because high-norm examples, prioritized for variance reduction, are often adversarial outliers. We formalize this misalignment using an $\varepsilon$-contamination model and propose Disagreement-Regularized Importance Sampling (DR-IS), a sub-sampling method based on loss rank-disagreement across independent proxy ensemble. We prove finite-sample concentration bounds showing that the empirical rank disagreement of bulk corrupted examples is bounded above, and that of boundary-clean examples bounded below, both at rate $O(\sqrt{\log(N/δ)/K})$ with probability $1-δ$; when the structural expectation gap $Δ'$ between the two groups is positive and the boundary-clean set is at least as large as the selected subset, these bounds certify strict separation and control the contamination rate of the selected subset. Empirically, DR-IS remains robust under targeted high-norm attacks that break magnitude-based methods such as the Error $L_2$-norm (EL2N) on benchmark datasets. DR-IS complements training-dynamics approaches like Area Under the Margin ranking (AUM), offering improved robustness in the loss-aligned regime alongside explicit finite-sample concentration certificates and a contamination bound limiting noise leakage from the statistical tail of corrupted points.
format Preprint
id arxiv_https___arxiv_org_abs_2605_07551
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Disagreement-Regularized Importance Sampling for Adversarial Label Corruption
Horváth, Csongor
Sintorn, Ida-Maria
Singh, Prashant
Machine Learning
Standard Importance Sampling (IS) collapses under label corruption because high-norm examples, prioritized for variance reduction, are often adversarial outliers. We formalize this misalignment using an $\varepsilon$-contamination model and propose Disagreement-Regularized Importance Sampling (DR-IS), a sub-sampling method based on loss rank-disagreement across independent proxy ensemble. We prove finite-sample concentration bounds showing that the empirical rank disagreement of bulk corrupted examples is bounded above, and that of boundary-clean examples bounded below, both at rate $O(\sqrt{\log(N/δ)/K})$ with probability $1-δ$; when the structural expectation gap $Δ'$ between the two groups is positive and the boundary-clean set is at least as large as the selected subset, these bounds certify strict separation and control the contamination rate of the selected subset. Empirically, DR-IS remains robust under targeted high-norm attacks that break magnitude-based methods such as the Error $L_2$-norm (EL2N) on benchmark datasets. DR-IS complements training-dynamics approaches like Area Under the Margin ranking (AUM), offering improved robustness in the loss-aligned regime alongside explicit finite-sample concentration certificates and a contamination bound limiting noise leakage from the statistical tail of corrupted points.
title Disagreement-Regularized Importance Sampling for Adversarial Label Corruption
topic Machine Learning
url https://arxiv.org/abs/2605.07551