Enhancing Conformal Prediction Using E-Test Statistics
This work addresses uncertainty quantification for machine learning practitioners, but it appears incremental as it modifies an existing framework without claiming broad breakthroughs.
The paper tackles the problem of uncertainty quantification in machine learning predictions by proposing an alternative approach to conformal prediction that uses e-test statistics instead of p-values, resulting in the introduction of a bounded-below predictor to enhance efficacy.
Conformal Prediction (CP) serves as a robust framework that quantifies uncertainty in predictions made by Machine Learning (ML) models. Unlike traditional point predictors, CP generates statistically valid prediction regions, also known as prediction intervals, based on the assumption of data exchangeability. Typically, the construction of conformal predictions hinges on p-values. This paper, however, ventures down an alternative path, harnessing the power of e-test statistics to augment the efficacy of conformal predictions by introducing a BB-predictor (bounded from the below predictor).