A Refined Analysis of UCBVI
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918037905473536 |
|---|---|
| author | Drago, Simone Mussi, Marco Metelli, Alberto Maria |
| author_facet | Drago, Simone Mussi, Marco Metelli, Alberto Maria |
| contents | In this work, we provide a refined analysis of the UCBVI algorithm (Azar et al., 2017), improving both the bonus terms and the regret analysis. Additionally, we compare our version of UCBVI with both its original version and the state-of-the-art MVP algorithm. Our empirical validation demonstrates that improving the multiplicative constants in the bounds has significant positive effects on the empirical performance of the algorithms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_17370 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Refined Analysis of UCBVI Drago, Simone Mussi, Marco Metelli, Alberto Maria Machine Learning In this work, we provide a refined analysis of the UCBVI algorithm (Azar et al., 2017), improving both the bonus terms and the regret analysis. Additionally, we compare our version of UCBVI with both its original version and the state-of-the-art MVP algorithm. Our empirical validation demonstrates that improving the multiplicative constants in the bounds has significant positive effects on the empirical performance of the algorithms. |
| title | A Refined Analysis of UCBVI |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2502.17370 |