Abstract
This paper discusses a counterpart of conformal prediction for e-values, conformal e-prediction. Conformal e-prediction is conceptually simpler and had been developed in the 1990s as a precursor of conformal prediction. When conformal prediction emerged as result of replacing e-values by p-values, it seemed to have important advantages over conformal e-prediction without obvious disadvantages. This paper re-examines relations between conformal prediction and conformal e-prediction systematically from a modern perspective. Conformal e-prediction has advantages of its own, such as the ease of designing conditional conformal e-predictors and the guaranteed validity of cross-conformal e-predictors (whereas for cross-conformal predictors validity is only an empirical fact and can be broken with excessive randomization). Even where conformal prediction has clear advantages, conformal e-prediction can often emulate those advantages, more or less successfully.
| Original language | English |
|---|---|
| Article number | 111674 |
| Number of pages | 11 |
| Journal | Pattern Recognition |
| Volume | 166 |
| Early online date | 17 Apr 2025 |
| DOIs | |
| Publication status | Published - Oct 2025 |
Keywords
- conformal e-prediction
- cross-conformal e-prediction
- e-value
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver