Aaditya Ramdas

Uncertainty quantification using martingales for misspecified Gaussian processes

2020 · arXiv (Cornell University)

We address uncertainty quantification for Gaussian processes (GPs) under misspecified priors, with an eye towards Bayesian Optimization (BO). GPs are widely used in BO because they easily enable exploration based on posterior uncertainty bands. However, …

Distribution-free calibration guarantees for histogram binning without sample splitting

2021 · arXiv (Cornell University)

We prove calibration guarantees for the popular histogram binning (also called uniform-mass binning) method of Zadrozny and Elkan [2001]. Histogram binning has displayed strong practical performance, but theoretical guarantees have only been shown for sample …

Deep anytime-valid hypothesis testing

2023 · arXiv (Cornell University)

We propose a general framework for constructing powerful, sequential hypothesis tests for a large class of nonparametric testing problems. The null hypothesis for these problems is defined in an abstract form using the action of …