TY - GEN
T1 - ADOM: Accelerated Decentralized Optimization Method for Time-Varying Networks
AU - Kovalev, Dmitry
AU - Shulgin, Egor
AU - Richtarik, Peter
AU - Rogozin, Alexander
AU - Gasnikov, Alexander
N1 - KAUST Repository Item: Exported on 2021-11-18
Acknowledgements: The work of D. Kovalev, E. Shulgin and P. Richt ' arik was supported by the KAUST Baseline Research Funding Scheme. The work of A. Rogozin and A. Gasnikov was supported by the Russian Science Foundation (project 21-71-30005).
PY - 2021
Y1 - 2021
N2 - We propose ADOM - an accelerated method for smooth and strongly convex decentralized optimization over time-varying networks. ADOM uses a dual oracle, i.e., we assume access to the gradient of the Fenchel conjugate of the individual loss functions. Up to a constant factor, which depends on the network structure only, its communication complexity is the same as that of accelerated Nesterov gradient method (Nesterov, 2003). To the best of our knowledge, only the algorithm of Rogozin et al. (2019) has a convergence rate with similar properties. However, their algorithm converges under the very restrictive assumption that the number of network changes can not be greater than a tiny percentage of the number of iterations. This assumption is hard to satisfy in practice, as the network topology changes usually can not be controlled. In contrast, ADOM merely requires the network to stay connected throughout time.
AB - We propose ADOM - an accelerated method for smooth and strongly convex decentralized optimization over time-varying networks. ADOM uses a dual oracle, i.e., we assume access to the gradient of the Fenchel conjugate of the individual loss functions. Up to a constant factor, which depends on the network structure only, its communication complexity is the same as that of accelerated Nesterov gradient method (Nesterov, 2003). To the best of our knowledge, only the algorithm of Rogozin et al. (2019) has a convergence rate with similar properties. However, their algorithm converges under the very restrictive assumption that the number of network changes can not be greater than a tiny percentage of the number of iterations. This assumption is hard to satisfy in practice, as the network topology changes usually can not be controlled. In contrast, ADOM merely requires the network to stay connected throughout time.
UR - http://hdl.handle.net/10754/667661
UR - https://arxiv.org/pdf/2102.09234.pdf
M3 - Conference contribution
BT - International Conference on Machine Learning (ICML)
PB - arXiv
ER -