A revised arXiv paper proposes a noise-adaptive approach to conformal inference for classification. Standard conformal methods give prediction sets with guaranteed coverage, but the work addresses how that reliability holds when label or feature noise is present. The method targets marginal coverage guarantees while adjusting to noisy data conditions.
A new arXiv paper proposes a conformal prediction approach that draws on multiple heterogeneous data sources rather than treating them as one pool. The method exploits differences between sources through localization, aiming to keep prediction sets reliable when the test distribution departs from any single source. This targets settings where combining sources is useful but naive pooling would break coverage guarantees.
A new arXiv paper addresses a common problem in medical image classification: transitional categories whose features overlap with neighboring classes, creating fuzzy decision boundaries. The proposed approach, adaptive conformal redistribution, aims to sharpen uncertainty-aware prediction sets so they remain informative when a case sits between two diagnoses. Conformal prediction normally provides coverage guarantees for such sets, and the work adapts that framework to inter-class transitions.
A new arXiv paper relaxes the standard no-interference assumption used in conformal counterfactual prediction, where one unit's treatment is assumed not to affect another's outcome. The authors develop an approach that produces prediction sets with finite-sample coverage guarantees for counterfactual outcomes and individual treatment effects when units interact within a network. This matters for settings such as social networks, marketplaces, and trials where spillover effects are common.
A new arXiv paper introduces SafeImpute, a method for filling in missing laboratory values in clinical datasets where patient visits are irregular and tests are ordered unevenly. The approach applies conformal selection to provide reliability guarantees, rather than only improving average imputation accuracy. It aims to give clinicians more dependable guidance when key lab indicators are absent.
A revised arXiv preprint surveys and compares non-conformity score functions used in conformal prediction for classification. Conformal prediction generates prediction sets rather than single labels, with guarantees that the sets cover the true class at a chosen rate. The work evaluates how different scoring choices affect the efficiency and validity of those sets.
A new arXiv paper frames novel-view synthesis with 3D Gaussian Splatting as a structured regression problem and applies conformal prediction to it. The goal is to give rendered views formal statistical coverage guarantees, going beyond uncertainty heatmaps that offer no such certification. The approach is aimed at settings where rendered outputs must meet verifiable reliability standards.
A new arXiv paper introduces CertDW, a method for verifying whether a public dataset was used to train a given deep neural network. The approach applies conformal calibration to produce certified ownership decisions, addressing weaknesses the authors identify in existing dataset ownership verification techniques. The work targets copyright protection for open-source datasets such as ImageNet.