Anytime-Valid Generalized Universal Inference on Risk Minimizers

3299 

Contributed Abstract 

Poster 

Neil Dey (1), Ryan Martin (1), Jonathan Williams (1)

(1) North Carolina State University, Raleigh, North Carolina

Ryan Martin  
North Carolina State University

Jonathan Williams  
North Carolina State University

Neil Dey  
North Carolina State University

Neil Dey  
N/A

A common goal in statistics and machine learning is estimation of unknowns. Point estimates alone are of little value without an accompanying measure of uncertainty, but traditional uncertainty quantification methods, such as confidence sets and p-values, often require strong distributional or structural assumptions that may not be justified in modern problems. The present paper considers a very common case in machine learning, where the quantity of interest is the minimizer of a given risk (expected loss) function. For such cases, we propose a generalized universal procedure for inference on risk minimizers that features a finite-sample, frequentist validity property under mild distributional assumptions. One version of the proposed procedure is shown to be anytime-valid in the sense that it maintains validity properties regardless of the stopping rule used for the data collection process. We show how this anytime-validity property offers protection against certain factors contributing to the replication crisis in science.

e-process|e-value|empirical risk minimization|Gibbs posterior|learning rate|machine learning

Uncertainty Quantification in Complex Systems Interest Group

Miscellaneous