Papers › Graph topology invariant gradient and sampling complexity for decentralized and...
Graph topology invariant gradient and sampling complexity for decentralized and stochastic optimization
Guanghui Lan, Yuyuan Ouyang, Yi Zhou
The archive published only this paper's code-link row. Authors, date and abstract are from arXiv's metadata (CC0), read from the Kaggle arXiv metadata snapshot of 2026-09-12 where its title matched the archive's; the title is the archive's.
One fundamental problem in decentralized multi-agent optimization is the trade-off between gradient/sampling complexity and communication complexity. We propose new algorithms whose gradient and sampling complexities are graph topology invariant while their communication complexities remain optimal. For convex smooth deterministic problems, we propose a primal dual sliding (PDS) algorithm that computes an ϵ-solution with O((L̃/ϵ)^(1/2)) gradient and O((L̃/ϵ)^(1/2)+‖𝒜‖/ϵ) communication complexities, where L̃ is the smoothness parameter of the objective and 𝒜 is related to either the graph Laplacian or the transpose of the oriented incidence matrix of the communication network. The results can be improved to O((L̃/μ)^(1/2)log(1/ϵ)) and O((L̃/μ)^(1/2)log(1/ϵ) + ‖𝒜‖/ϵ^(1/2)) respectively with μ-strong convexity. We also propose a stochastic variant, the primal dual sliding (SPDS) algorithm for problems with stochastic gradients. The SPDS algorithm utilizes the mini-batch technique and enables the agents to perform sampling and communication simultaneously. It computes a stochastic ϵ-solution with O((L̃/ϵ)^(1/2) + (σ/ϵ)²) sampling complexity, which can be improved to O((L̃/μ)^(1/2)log(1/ϵ) + σ²/ϵ) with strong convexity. Here σ² is the variance. The communication complexities of SPDS remain the same as that of the deterministic case. All the aforementioned gradient and sampling complexities match the lower complexity bounds for centralized convex smooth optimization and are independent of the network structure. To the best of our knowledge, these gradient and sampling complexities have not been obtained before for decentralized optimization over a constraint feasible set.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Results from the paper archive 2025-07-28
No leaderboard rows for this paper in the archive.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections