Two novel online conformal prediction algorithms enforce nested prediction sets across coverage levels using online optimization with regret bounds for quantile error control.
Advances in Neural Information Processing Systems , volume=
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
years
2026 8verdicts
UNVERDICTED 8representative citing papers
SeqRejectron constructs a stopping rule with a small set of validator policies to achieve horizon-free sample complexity for selective imitation learning under arbitrary dynamics shifts.
HAPS constructs shorter conformal prediction sets for censored time-to-event outcomes by using time-varying covariate histories and IPCW, achieving approximate coverage among survivors with up to 75% shorter intervals in simulations.
Risk-controlled post-processing yields a threshold-structured policy that follows the baseline except where an oracle fallback sharply reduces conditional violation risk, achieving O(log n/n) expected excess risk in i.i.d. settings and exact risk control under exchangeability.
Trimming helps conformal prediction under contamination precisely when the anomaly score separates retention probabilities without biasing clean scores, otherwise the retained mixture coefficient prevents substantial decontamination.
SA-BCP adaptively blends temporal Bayesian predictions with spatial KDE evidence via threshold K, derives closed-form MSE-optimal K, and provides an online selection procedure with regret bounds, yielding sharper intervals at nominal coverage on volatility and weather data.
STOIC integrates STGNN point forecasting with tabular foundation model in-context learning for conformal prediction to quantify uncertainty in graph-structured energy time series.
A model-agnostic two-stage estimator for conditional quantiles that represents the high-fidelity quantile as a low-fidelity quantile evaluated at a covariate-dependent level, with theory on faster convergence rates under shape similarity.
citing papers explorer
-
Online Conformal Prediction: Enforcing monotonicity via Online Optimization
Two novel online conformal prediction algorithms enforce nested prediction sets across coverage levels using online optimization with regret bounds for quantile error control.
-
Learning When to Stop: Selective Imitation Learning Under Arbitrary Dynamics Shift
SeqRejectron constructs a stopping rule with a small set of validator policies to achieve horizon-free sample complexity for selective imitation learning under arbitrary dynamics shifts.
-
History-Aware Conformal Prediction Sets for Censored Time-to-Event Outcomes
HAPS constructs shorter conformal prediction sets for censored time-to-event outcomes by using time-varying covariate histories and IPCW, achieving approximate coverage among survivors with up to 75% shorter intervals in simulations.
-
Risk-Controlled Post-Processing of Decision Policies
Risk-controlled post-processing yields a threshold-structured policy that follows the baseline except where an oracle fallback sharply reduces conditional violation risk, achieving O(log n/n) expected excess risk in i.i.d. settings and exact risk control under exchangeability.
-
When Does Trimming Help Conformal Prediction? A Retained-Law Diagnostic under Calibration Contamination
Trimming helps conformal prediction under contamination precisely when the anomaly score separates retention probabilities without biasing clean scores, otherwise the retained mixture coefficient prevents substantial decontamination.
-
Optimal Spatio-Temporal Decoupling for Bayesian Conformal Prediction
SA-BCP adaptively blends temporal Bayesian predictions with spatial KDE evidence via threshold K, derives closed-form MSE-optimal K, and provides an online selection procedure with regret bounds, yielding sharper intervals at nominal coverage on volatility and weather data.
-
Relational and Sequential Conformal Inference for Energy Time Series over Graphs via Foundation Models
STOIC integrates STGNN point forecasting with tabular foundation model in-context learning for conformal prediction to quantify uncertainty in graph-structured energy time series.
-
Multi-Fidelity Quantile Regression
A model-agnostic two-stage estimator for conditional quantiles that represents the high-fidelity quantile as a low-fidelity quantile evaluated at a covariate-dependent level, with theory on faster convergence rates under shape similarity.