Adaptive Consensus ADMM for Distributed Optimization
By Zheng Xu, Gavin Taylor, Hao Li, Mário A. T. Figueiredo, Xiaoming Yuan, Tom Goldstein. 2017
73 citations recorded by the source corpus.
The alternating direction method of multipliers (ADMM) is commonly used for distributed model fitting problems, but its performance and reliability depend strongly on user-defined penalty parameters. We study distributed ADMM methods that boost performance by using different fine-tuned algorithm parameters on each worker node. We present a O(1/k) convergence rate for adaptive ADMM methods with node-specific parameters, and propose adaptive consensus ADMM (ACADMM), which automatically tunes parameters without user oversight.
Subfields: Artificial Intelligence, Computational Mechanics
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Ji Liu
- Hanlin Tang
- Xiangru Lian
- Tong Zhang
- Sebastian U. Stich
- Martin Jaggi
- Peter Richtárik
- Xueru Zhang
In Computational Mechanics
- Ji Liu
- Hanlin Tang
- Tong Zhang
- Xiangru Lian
- Peter Richtárik
- Xueru Zhang
- Mohammad Mahdi Khalili
- M. Liu
Citations in the corpus
- DoubleSqueeze: Parallel Stochastic Gradient Descent with Double-Pass Error-Compensated Compression (2019, methodology): “There has been lots of works studying the implementation of distributed learning from different angles. such as differentially private distributed optimization (Jayaraman et al., 2018; Zhang et al., 2018), adaptive distributed ADMM (Xu et al., 2017), adaptive distributed SGD (Cutkosky & Busa-Fekete,”
- Distributed Learning over Unreliable Networks (2018, neutral): “In Xu et al. (2017), an adaptive consensus ADMM is proposed and Goldstein et al. (2016) studied the performance of transpose ADMM on an entire distributed dataset.”
- Recycled ADMM: Improving the Privacy and Accuracy of Distributed Algorithms (2019, background): “Existing approaches to decentralizing the above problem primarily consist of subgradient-based algorithms [7]–[9] and ADMM-based algorithms [10]–[16].”