Composite Optimization with Coupling Constraints via Dual Proximal Gradient Method with Applications to Asynchronous Networks
arXiv:2102.12797
Abstract
In this paper, we consider solving a composite optimization problem with coupling constraints in a multi-agent network based on proximal gradient method. In this problem, all the agents jointly minimize the sum of individual cost functions composed of smooth and possibly non-smooth parts. To this end, we derive the dual problem by the concept of Fenchel conjugate, which gives rise to the dual proximal gradient algorithm by allowing for the asymmetric individual interpretations of the global constraints. Then, an asynchronous dual proximal gradient algorithm is proposed for the asynchronous networks with heterogenous step-sizes and communication delays. For both the two algorithms, if the non-smooth parts of the objective functions are simple-structured, we only need to update dual variables by some simple operations, accounting for the reduction of the overall computational complexity. Analytical convergence rate of the proposed algorithms is derived and their efficacy is verified by solving a social welfare optimization problem of electricity market in the numerical simulation.
arXiv comment18 pages, 5 figures
References in corpus (10)
- Speeding Up Distributed Machine Learning Using Codes
- Multi-Agent Distributed Optimization via Inexact Consensus ADMM
- Distributed Constrained Optimization by Consensus-Based Primal-Dual Perturbation Method
- Asynchronous Distributed ADMM for Large-Scale Optimization- Part I: Algorithm and Convergence Analysis
- A Novel Consensus-based Distributed Algorithm for Economic Dispatch Based on Local Estimation of Power Mismatch
- Bandit Convex Optimization for Scalable and Dynamic IoT Management
- A Proximal Dual Consensus ADMM Method for Multi-Agent Constrained Optimization
- A Generalized Accelerated Composite Gradient Method: Uniting Nesterov's Fast Gradient Method and FISTA
- Distributed Proximal Gradient Algorithm for Partially Asynchronous Computer Clusters
- SCOPE: Scalable Composite Optimization for Learning on Spark