AI Paper Atlas

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

CAP: A General Algorithm for Online Selective Conformal Prediction with FCR Control

Yajie Bao, Yuyang Huo, Haojie Ren, Changliang Zou

JMLR 2025 · Learning theory and algorithms › Prediction and decision losses · approach: Conformal prediction sets · 12 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 CAP adaptive pick calibration instead of online ordinary conformal prediction (OCP)online selective conformal prediction · synthetic Scenarios A-C, five selection rulesreal-time FCRImproved
use CAP instead of LORD-CIonline selective conformal prediction · synthetic Scenarios A-CPI lengthImproved
add DtACI-style level adaptation with decaying rates (CAP-DtACI) to CAP (ablation)CAP · Scenario B under slow shift, change point and time series shiftreal-time FCRImproved
add selection-aware calibration (CAP) to DtACI (ablation)DtACI · Scenario B under four distribution shiftsreal-time FCRImproved
CAP without level adaptation under distribution shift (ablation)CAP · Scenario B shifts, NVIDIA volatilityreal-time FCRHurt
use CAP instead of OCP and LORD-CIonline selective conformal prediction · DAVIS drug discoveryFCR and PI lengthImproved
vary initial holdout size n (ablation)CAP · syntheticFCRNo clear effect
use e-LOND-CI instead of CAPonline selective inference · syntheticFCR and PI lengthHurt

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