Cooperative Online Learning with Feedback Graphs
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2021
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910561823883264 |
|---|---|
| author | Cesa-Bianchi, Nicolò Cesari, Tommaso R. Della Vecchia, Riccardo |
| author_facet | Cesa-Bianchi, Nicolò Cesari, Tommaso R. Della Vecchia, Riccardo |
| contents | We study the interplay between communication and feedback in a cooperative online learning setting, where a network of communicating agents learn a common sequential decision-making task through a feedback graph. We bound the network regret in terms of the independence number of the strong product between the communication network and the feedback graph. Our analysis recovers as special cases many previously known bounds for cooperative online learning with expert or bandit feedback. We also prove an instance-based lower bound, demonstrating that our positive results are not improvable except in pathological cases. Experiments on synthetic data confirm our theoretical findings. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2106_04982 |
| institution | arXiv |
| publishDate | 2021 |
| record_format | arxiv |
| spellingShingle | Cooperative Online Learning with Feedback Graphs Cesa-Bianchi, Nicolò Cesari, Tommaso R. Della Vecchia, Riccardo Machine Learning We study the interplay between communication and feedback in a cooperative online learning setting, where a network of communicating agents learn a common sequential decision-making task through a feedback graph. We bound the network regret in terms of the independence number of the strong product between the communication network and the feedback graph. Our analysis recovers as special cases many previously known bounds for cooperative online learning with expert or bandit feedback. We also prove an instance-based lower bound, demonstrating that our positive results are not improvable except in pathological cases. Experiments on synthetic data confirm our theoretical findings. |
| title | Cooperative Online Learning with Feedback Graphs |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2106.04982 |