Distributed Conditional Gradient Algorithm for Two-Network Saddle-point Problem

We consider a two-network saddle-point problem with constraints, whose projections are expensive. We propose a projection-free algorithm, which is referred to as Distributed Frank-Wolfe Saddle-Point algorithm (DFWSP), which combines the gradient tracking technique and Frank-Wolfe technique. We prove...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Chinese Control and Decision Conference s. 4261 - 4266
Hlavní autoři: Hou, Jie, Zeng, Xianlin
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 15.08.2022
Témata:
Convergence
Convergence Rate
Distributed Algorithm
Distributed computing
Optimization
Projection-Free Method
Saddle-Point Problem
ISSN:1948-9447
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:We consider a two-network saddle-point problem with constraints, whose projections are expensive. We propose a projection-free algorithm, which is referred to as Distributed Frank-Wolfe Saddle-Point algorithm (DFWSP), which combines the gradient tracking technique and Frank-Wolfe technique. We prove that the algorithm achieves O(1/k 2 ) convergence rate for strongly-convex-strongly-concave saddle-point problems. We empirically shows that the proposed algorithm has better numerical performance than the distributed projected saddle-point algorithm.
ISSN:1948-9447
DOI:10.1109/CCDC55256.2022.10033870