Innovation Compression for Communication-Efficient Distributed Optimization With Linear Convergence
IEEE Transactions on Automatic Control · 2023-11 · 68 (11) · 6899-6906
DOI: 10.1109/tac.2023.3241771 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Automatic Control, 2023-11. Verfasst von Jiaqi Zhang und 2 weiteren Autor:innen. Die Arbeit wurde bisher 30-mal zitiert (Crossref) und nennt 29 Literaturangaben. Gefördert von: Ministry of Science and Technology of China, National Natural Science Foundation of China, Tsinghua University Initiative Scientific Research.
Abstract (Originalsprache)
Information compression is essential to reduce communication cost in distributed optimization over peer-to-peer networks. This article proposes a communication-efficient linearly convergent distributed (COLD) algorithm to solve strongly convex optimization problems. By compressing innovation vectors, which are the differences between decision vectors and their estimates, COLD achieves linear convergence for a class of $\delta$-contracted compressors, and we explicitly quantify how the compression affects the convergence rate. Interestingly, our results strictly improve existing results for the quantized consensus problem. Numerical experiments demonstrate the advantages of COLD under different compressors.
Förderung
- Ministry of Science and Technology of China
- National Natural Science Foundation of China
- Tsinghua University Initiative Scientific Research
Zitieren
APA
Zhang, J., You, K., & Xie, L. (2023). Innovation Compression for Communication-Efficient Distributed Optimization With Linear Convergence. IEEE Transactions on Automatic Control, 68(11), 6899-6906. https://doi.org/10.1109/tac.2023.3241771
Vancouver
Zhang J, You K, Xie L. Innovation Compression for Communication-Efficient Distributed Optimization With Linear Convergence. IEEE Transactions on Automatic Control. 2023;68(11):6899-6906. doi:10.1109/tac.2023.3241771
BibTeX
@article{Zhang2023,
author = {Zhang, Jiaqi and You, Keyou and Xie, Lihua},
title = {Innovation Compression for Communication-Efficient Distributed Optimization With Linear Convergence},
journal = {IEEE Transactions on Automatic Control},
year = {2023},
volume = {68},
number = {11},
pages = {6899--6906},
doi = {10.1109/tac.2023.3241771}
}
EUVIMED