8 papers
Robust Bayes-Assisted Conformal Prediction
Kianoosh Ashouritaklimi, Stefano Cortinovis, François Caron
Bayes-assisted conformal prediction combines the strengths of Bayesian modelling with exact, distribution-free frequentist coverage guarantees. Although conformal validity is prese…
Variational predictive resampling
Laura Battaglia, Stefano Cortinovis, Chris Holmes +2
Bayesian inference provides principled uncertainty quantification, but accurate posterior sampling with MCMC can be computationally prohibitive for modern applications. Variational…
Asymptotically Log-Optimal Bayes-Assisted Confidence Sequences for Bounded Means
Valentin Kilian, Stefano Cortinovis, François Caron
Confidence sequences based on test martingales provide time-uniform uncertainty quantification for the mean of bounded IID observations without parametric distributional assumption…
Inverse-Free Sparse Variational Gaussian Processes
Stefano Cortinovis, Laurence Aitchison, Stefanos Eleftheriadis +1
Gaussian processes (GPs) offer appealing properties but are costly to train at scale. Sparse variational GP (SVGP) approximations reduce cost yet still rely on Cholesky decompositi…
Confidence sequences with informative, bounded-influence priors
Stefano Cortinovis, Valentin Kilian, François Caron
Confidence sequences are collections of confidence regions that simultaneously cover the true parameter for every sample size at a prescribed confidence level. Tightening these seq…
Anytime-valid, Bayes-assisted, Prediction-Powered Inference
Valentin Kilian, Stefano Cortinovis, François Caron
Given a large pool of unlabelled data and a smaller amount of labels, prediction-powered inference (PPI) leverages machine learning predictions to increase the statistical efficien…
FAB-PPI: Frequentist, Assisted by Bayes, Prediction-Powered Inference
Stefano Cortinovis, François Caron
Prediction-powered inference (PPI) enables valid statistical inference by combining experimental data with machine learning predictions. When a sufficient number of high-quality pr…
Bayes-assisted Confidence Regions: Focal Point Estimator and Bounded-influence Priors
Stefano Cortinovis, François Caron
The Frequentist, Assisted by Bayes (FAB) framework constructs confidence regions that leverage prior information about parameter values. FAB confidence regions (FAB-CRs) have small…