A decentralised forward-backward-type algorithm with network-independent heterogenous agent step sizes
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908709171494912 |
|---|---|
| author | Tam, Matthew K Timms, Liam Zhang, Lele |
| author_facet | Tam, Matthew K Timms, Liam Zhang, Lele |
| contents | Consider the problem of finding a zero of a finite sum of maximally monotone operators, where some operators are Lipschitz continuous and the rest are potentially set-valued. We propose a forward-backward-type algorithm for this problem suitable for decentralised implementation. In each iteration, agents evaluate a Lipschitz continuous operator and the resolvent of a potentially set-valued operator, and then communicate with neighbouring agents. Agents choose their step sizes independently using only local information, and the step size upper bound has no dependence on the communication graph. We demonstrate the potential advantages of the proposed algorithm with numerical results for min-max problems and aggregative games. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_12502 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A decentralised forward-backward-type algorithm with network-independent heterogenous agent step sizes Tam, Matthew K Timms, Liam Zhang, Lele Optimization and Control 47N10, 65K15, 47H05, 47J22, 91A10 Consider the problem of finding a zero of a finite sum of maximally monotone operators, where some operators are Lipschitz continuous and the rest are potentially set-valued. We propose a forward-backward-type algorithm for this problem suitable for decentralised implementation. In each iteration, agents evaluate a Lipschitz continuous operator and the resolvent of a potentially set-valued operator, and then communicate with neighbouring agents. Agents choose their step sizes independently using only local information, and the step size upper bound has no dependence on the communication graph. We demonstrate the potential advantages of the proposed algorithm with numerical results for min-max problems and aggregative games. |
| title | A decentralised forward-backward-type algorithm with network-independent heterogenous agent step sizes |
| topic | Optimization and Control 47N10, 65K15, 47H05, 47J22, 91A10 |
| url | https://arxiv.org/abs/2512.12502 |