Small steps no more: Global convergence of stochastic gradient bandits for arbitrary learning rates
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915145654992896 |
|---|---|
| author | Mei, Jincheng Dai, Bo Agarwal, Alekh Vaswani, Sharan Raj, Anant Szepesvari, Csaba Schuurmans, Dale |
| author_facet | Mei, Jincheng Dai, Bo Agarwal, Alekh Vaswani, Sharan Raj, Anant Szepesvari, Csaba Schuurmans, Dale |
| contents | We provide a new understanding of the stochastic gradient bandit algorithm by showing that it converges to a globally optimal policy almost surely using \emph{any} constant learning rate. This result demonstrates that the stochastic gradient algorithm continues to balance exploration and exploitation appropriately even in scenarios where standard smoothness and noise control assumptions break down. The proofs are based on novel findings about action sampling rates and the relationship between cumulative progress and noise, and extend the current understanding of how simple stochastic gradient methods behave in bandit settings. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_07141 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Small steps no more: Global convergence of stochastic gradient bandits for arbitrary learning rates Mei, Jincheng Dai, Bo Agarwal, Alekh Vaswani, Sharan Raj, Anant Szepesvari, Csaba Schuurmans, Dale Machine Learning We provide a new understanding of the stochastic gradient bandit algorithm by showing that it converges to a globally optimal policy almost surely using \emph{any} constant learning rate. This result demonstrates that the stochastic gradient algorithm continues to balance exploration and exploitation appropriately even in scenarios where standard smoothness and noise control assumptions break down. The proofs are based on novel findings about action sampling rates and the relationship between cumulative progress and noise, and extend the current understanding of how simple stochastic gradient methods behave in bandit settings. |
| title | Small steps no more: Global convergence of stochastic gradient bandits for arbitrary learning rates |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2502.07141 |