Optimization of Distributed ADMM Algorithm Based on Minimum Network Latency

In order to solve the problem that the running speed of distributed ADMM algorithm is limited by the nodes with large network latency, a master node selection policy is proposed, which is based on the communication delay between network nodes, and the policy aims to reduce the running time of distri...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Proceedings of the International Symposium on Computational Intelligence and Design S. 7 - 10
Hauptverfasser: Zeng, Shuai, Wang, Shuangsheng, Zhang, Yi
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.12.2019
Schlagworte:
ADMM
algorithm optimizartion
Clustering algorithms
Convergence
Delays
Linear programming
machine learning
Machine learning algorithms
network latency
Optimization
Signal processing algorithms
ISSN:2473-3547
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In order to solve the problem that the running speed of distributed ADMM algorithm is limited by the nodes with large network latency, a master node selection policy is proposed, which is based on the communication delay between network nodes, and the policy aims to reduce the running time of distributed ADMM algorithm in the connected network. Experiment results show that the policy can effectively reduce the running time of distributed ADMM algorithm in connected networks, and the calculation accuracy is not affected. And the smaller the connection probability between network nodes, the more significant the advantages of the proposed policy are. And when there is a communication failure between a few nodes in the connected network, the strategy proposed in this paper is not affected much.
ISSN:2473-3547
DOI:10.1109/ISCID.2019.10084