Dimension-free comparison estimates for suprema of some canonical processes
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913571403726848 |
|---|---|
| author | Sharma, Shivam |
| author_facet | Sharma, Shivam |
| contents | We obtain error approximation bounds between expected suprema of canonical processes that are generated by random vectors with independent coordinates and expected suprema of Gaussian processes. In particular, we obtain a sharper proximity estimate for Rademacher and Gaussian complexities. Our estimates are dimension-free, and depend only on the geometric parameters and the numerical complexity of the underlying index set. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_14308 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Dimension-free comparison estimates for suprema of some canonical processes Sharma, Shivam Probability Statistics Theory We obtain error approximation bounds between expected suprema of canonical processes that are generated by random vectors with independent coordinates and expected suprema of Gaussian processes. In particular, we obtain a sharper proximity estimate for Rademacher and Gaussian complexities. Our estimates are dimension-free, and depend only on the geometric parameters and the numerical complexity of the underlying index set. |
| title | Dimension-free comparison estimates for suprema of some canonical processes |
| topic | Probability Statistics Theory |
| url | https://arxiv.org/abs/2312.14308 |