Saved in:
| Main Author: | |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2303.02770 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910793072640000 |
|---|---|
| author | F, Paulo C. Marques |
| author_facet | F, Paulo C. Marques |
| contents | When split conformal prediction operates in batch mode with exchangeable data, we determine the exact distribution of the empirical coverage of prediction sets produced for a finite batch of future observables, as well as the exact distribution of its almost sure limit when the batch size goes to infinity. Both distributions are universal, being determined solely by the nominal miscoverage level and the calibration sample size, thereby establishing a criterion for choosing the minimum required calibration sample size in applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2303_02770 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Universal distribution of the empirical coverage in split conformal prediction F, Paulo C. Marques Statistics Theory Machine Learning When split conformal prediction operates in batch mode with exchangeable data, we determine the exact distribution of the empirical coverage of prediction sets produced for a finite batch of future observables, as well as the exact distribution of its almost sure limit when the batch size goes to infinity. Both distributions are universal, being determined solely by the nominal miscoverage level and the calibration sample size, thereby establishing a criterion for choosing the minimum required calibration sample size in applications. |
| title | Universal distribution of the empirical coverage in split conformal prediction |
| topic | Statistics Theory Machine Learning |
| url | https://arxiv.org/abs/2303.02770 |