Xiaokai Chen, Ilya Kuruzov, Gesualdo Scutari
The paper introduces an adaptive decentralized optimization method using three-operator splitting and local stepsize adjustments, achieving robust convergence for convex and strongly convex problems.
This research addresses how to optimize complex functions across a network of agents, where each agent has its own local function to minimize. The authors propose a new method that allows each agent to adjust its optimization steps based on local conditions, using a technique called three-operator splitting. This approach ensures that the optimization process is efficient and converges reliably, even when the functions involved are complex or have non-smooth parts. The method has been tested and shows promising results in terms of speed and accuracy.
