AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Bulk-Calibrated Credal Ambiguity Sets: Fast, Tractable Decision Making under Out-of-Sample Contamination

Mengqi Chen, Thomas Berrett, Theodoros Damoulas, Michele Caprio

ICML 2026 · Learning theory and algorithms › Prediction and decision losses · approach: Distributionally robust optimization · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use LV bulk-calibrated contamination objective instead of Wasserstein DROlinear regression with absolute loss · California housing East->West shiftMAE (x10^4): 12.3 -> 10.7Improved
use LV objective instead of Wasserstein DROlinear regression with absolute loss · California housing East->West shiftCVaR2% (x10^4): 26.5 -> 23.9Improved
add LV contamination layer on top of GroupDRO (LV-Group)GroupDRO linear head · CivilCommentsworst-group accuracy: 0.456 -> 0.516Improved
add LV contamination layer on top of GroupDRO (LV-Group)GroupDRO linear head · CivilCommentsmean accuracy: 0.770 -> 0.828Improved
train GroupDRO on bulk-filtered data only (without the LV contamination term) (ablation)GroupDRO · CivilCommentsworst-group accuracy: 0.456 -> 0.453No clear effect
use LV-Group instead of ERMlinear head · CivilCommentsworst-group accuracy: 0.022 -> 0.516 (mean acc 0.891 -> 0.828)Improved
use LV objective with fewer SAA samples (truncated 0.5M) than KL baselinesnewsvendor Bayesian centre · Student-t newsvendortotal run time (s): 3.6 -> 1.3 vs KL-BASPPImproved
use LV under no test contamination with large epsilon_LVLV newsvendor · Student-t newsvendor (0% contamination)MSDHurt
use LV under contaminated test settings (10%, 20%)newsvendor · Student-t newsvendor (10% and 20% contamination)MSDImproved
increase KL tolerance beyond finite-scenario limit in KL-BDRO (ablation)KL-BDRO · Student-t newsvendorOOS mean-varianceNo clear effect
vary tail budget gamma around minimum certifiable value (ablation)LV newsvendor · Student-t newsvendorOOS performanceNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy