Stochastic distributed optimization under average second-order similarity: Algorithms and analysis

D Lin, Y Han, H Ye, Z Zhang - Advances in Neural …, 2024 - proceedings.neurips.cc
We study finite-sum distributed optimization problems involving a master node and $ n-1$
local nodes under the popular $\delta $-similarity and $\mu $-strong convexity conditions …

Smooth monotone stochastic variational inequalities and saddle point problems: A survey

A Beznosikov, B Polyak, E Gorbunov… - European Mathematical …, 2023 - ems.press
This paper is a survey of methods for solving smooth,(strongly) monotone stochastic
variational inequalities. To begin with, we present the deterministic foundation from which …

Two losses are better than one: Faster optimization using a cheaper proxy

B Woodworth, K Mishchenko… - … Conference on Machine …, 2023 - proceedings.mlr.press
We present an algorithm for minimizing an objective with hard-to-compute gradients by
using a related, easier-to-access function as a proxy. Our algorithm is based on approximate …

Method with batching for stochastic finite-sum variational inequalities in non-Euclidean setting

A Pichugin, M Pechin, A Beznosikov, V Novitskii… - Chaos, Solitons & …, 2024 - Elsevier
Variational inequalities are a universal optimization paradigm that incorporate classical
minimization and saddle point problems. Nowadays more and more tasks require to …

Faster federated optimization under second-order similarity

A Khaled, C ** - ar** efficient optimization algorithms, it is crucial to account for communication
constraints--a significant challenge in modern Federated Learning. The best-known …

Similarity, compression and local steps: three pillars of efficient communications for distributed variational inequalities

A Beznosikov, M Takác… - Advances in Neural …, 2024 - proceedings.neurips.cc
Variational inequalities are a broad and flexible class of problems that includes
minimization, saddle point, and fixed point problems as special cases. Therefore, variational …

Stochastic proximal point methods for monotone inclusions under expected similarity

A Sadiev, L Condat, P Richtárik - arxiv preprint arxiv:2405.14255, 2024 - arxiv.org
Monotone inclusions have a wide range of applications, including minimization, saddle-
point, and equilibria problems. We introduce new stochastic algorithms, with or without …

Local methods with adaptivity via scaling

S Chezhegov, S Skorik, N Khachaturov… - arxiv preprint arxiv …, 2024 - arxiv.org
The rapid development of machine learning and deep learning has introduced increasingly
complex optimization challenges that must be addressed. Indeed, training modern …

Accelerated Stochastic ExtraGradient: Mixing Hessian and gradient similarity to reduce communication in distributed and federated learning

D Bylinkin, K Degtyarev, A Beznosikov - arxiv preprint arxiv:2409.14280, 2024 - arxiv.org
Modern realities and trends in learning require more and more generalization ability of
models, which leads to an increase in both models and training sample size. It is already …