Adaptive Uncertainty-Weighted ADMM for Distributed Optimization

Autor:	Ye, Jianping, Wan, Caleb, Fung, Samy Wu
Rok vydání:	2021
Předmět:	Mathematics - Optimization and Control Mathematics - Numerical Analysis 65K10
Druh dokumentu:	Working Paper
Popis:	We present AUQ-ADMM, an adaptive uncertainty-weighted consensus ADMM method for solving large-scale convex optimization problems in a distributed manner. Our key contribution is a novel adaptive weighting scheme that empirically increases the progress made by consensus ADMM scheme and is attractive when using a large number of subproblems. The weights are related to the uncertainty associated with the solutions of each subproblem, and are efficiently computed using low-rank approximations. We show AUQ-ADMM provably converges and demonstrate its effectiveness on a series of machine learning applications, including elastic net regression, multinomial logistic regression, and support vector machines. We provide an implementation based on the PyTorch package. Comment: 16 pages, 10 figures
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2109.01089 Zobrazit plný text záznamu View this record from Arxiv