Distribution-Free, Risk-Controlling Prediction Sets

By Stephen Bates, Anastasios Nikolas Angelopoulos, Lihua Lei, Jitendra Malik, Michael I. Jordan. 2021

311 citations recorded by the source corpus.

While improving prediction accuracy has been the focus of machine learning in recent years, this alone does not suffice for reliable decision-making. Deploying learning systems in consequential settings also requires calibrating and communicating the uncertainty of predictions. To convey instance-wise uncertainty for prediction tasks, we show how to generate set-valued predictions from a black-box predictor that controls the expected loss on future test points at a user-specified level. Our approach provides explicit finite-sample guarantees for any dataset by using a holdout set to calibrate the size of the prediction sets. This framework enables simple, distribution-free, rigorous error control for many tasks, and we demonstrate it in five large-scale machine learning problems: (1) classification problems where some mistakes are more costly than others; (2) multi-label classification, where each observation has multiple associated labels; (3) classification problems where the labels have a hierarchical structure; (4) image segmentation, where we wish to predict a set of pixels containing an object of interest; and (5) protein structure prediction. Last, we discuss extensions to uncertainty quantification for ranking, metric learning, and distributionally robust learning.

Subfields: Artificial Intelligence, Statistical Learning Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistical Learning Theory

Citations in the corpus

Conformal Risk Control (2022, methodology): “This generalization of conformal prediction broadens its scope to new applications, as shown in Section 3.”
Few-shot Conformal Prediction with Auxiliary Tasks (2021, background): “…conformal prediction in the context of various validity conditions, such as conditional, risk-controlling, admissible, or equalized coverage (Chernozhukov et al., 2019; Cauchois et al., 2020; Kivaranovic et al., 2020; Romano et al., 2019; 2020; Bates et al., 2020; Fisch et al., 2021, inter alia ).”
Confident Adaptive Language Modeling (2022, methodology): “Our approach leverages recent techniques in distribution-free risk control in order to create conﬁdent generations with strong statistical guarantees [2; 3; 10].”