Read the original document by opening this link in a new tab.
Table of Contents
1. Introduction
2. Related Works
3. Preliminaries
3.1 Notations
3.2 Continuous DR-Submodularity
3.3 Problem Formulation
4. One-Shot Decentralized Meta-Frank-Wolfe Algorithm
Summary
The document presents two communication-efficient decentralized online algorithms for monotone continuous DR-submodular maximization. The first algorithm, Mono-DMFW, achieves a regret bound of O(T^4/5) and is the first one-shot and projection-free decentralized online algorithm for this problem. The second algorithm, DOBGA, attains a regret of O(p T) and is the first to achieve the optimal O(p T) regret against a (1-1/e)-approximation with only one gradient inquiry for each local objective function per step. Experimental results confirm the effectiveness of the proposed methods.