AI Paper Atlas

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

Multi-Distribution Robust Conformal Prediction

Yuqi YANG, Ying Jin

ICML 2026 · Learning theory and algorithms › Prediction and decision losses · approach: Conformal prediction sets · 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
learn a shared conformity score with the dual objective before max-p aggregation (MDCP vs naive max-p of single-source scores)max-p aggregation · Synthetic classification LINEARaverage set size: 34.39% smallerImproved
learn a shared conformity score before max-p aggregationmax-p aggregation · Synthetic regression LINEARaverage interval width: about 22% narrowerImproved
use max-p aggregation across sources instead of calibrating on a single sourcesingle-source conformal sets · synthetic and real data (FMoW, PovertyMap, MEPS)worst-case coverage: single-source coverage as low as about 0.6 on PovertyMapImproved
replace estimated densities with ground-truth densities (Oracle) (ablation)MDCP classification · Synthetic classificationaverage set sizeNo clear effect
replace estimated densities with ground-truth densities (Oracle) (ablation)MDCP regression · Synthetic regressioninterval widthImproved
add tie-breaking randomization U_k to p-values (ablation)MDCP · Appendix D.2coverage and sizeNo clear effect
add norm regularization with tuned penalty to the dual objective (ablation)MDCP · simulations and real datacoverage and sizeNo clear effect
increase inter-source separation (temperature tau) (ablation)single-source baselines vs MDCP · TEMPERATURE suitesworst-case coverageHurt

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