Benchmarking non-conformity score functions in conformal prediction
Sol Erika Boman · 2026
Read Benchmarking non-conformity score functions in conformal prediction by Sol Erika Boman (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-24983.txt.
Description
Sol Erika Boman benchmarks non-conformity scores for set-valued classification. It swaps single-class calls for prediction sets, guaranteeing the true class lands inside at or above a preset rate. How large and useful those sets are hinges on the score function chosen. The literature offers many score functions, but few studies compare their properties and power. Score properties, literature examples, and new modifications are given. A new method for judging set sizes enables comparison across functions, including class-conditional calls under imbalanced classes.
Catalog record
| CREATOR | Sol Erika Boman |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53934 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.24983