?
Breaking the Heavy-Tailed Noise Barrier in Stochastic Optimization Problems
P. 856–864.
We consider stochastic optimization problems with heavy-tailed noise with structured density. For such problems, we show that it is possible to get faster rates of convergence than 𝑂(𝐾^{−2(𝛼−1)/𝛼}), when the stochastic gradients have finite 𝛼-th moment, 𝛼∈(1,2]. In particular, our analysis allows the noise norm to have an unbounded expectation. To achieve these results, we stabilize stochastic gradients, using smoothed medians of means. We prove that the resulting estimates have negligible bias and controllable variance. This allows us to carefully incorporate them into clipped-SGD and clipped-SSTM and derive new high-probability complexity bounds in the considered setup.
Beznosikov A., Kormakov G., Grigorievskiy A. et al., Journal of Optimization Theory and Applications 2026 Vol. 209 Article 18
The objective of Vertical Federated Learning (VFL) is to collectively train a model using features available on different devices while sharing the same users. This paper focuses on the saddle point reformulation of the VFL problem via the classical Lagrangian function. We first demonstrate how this formulation can be solved using deterministic methods.More importantly, we explore various stochastic modifications to ...
Added: June 17, 2026
Финагин М. И., Экономический журнал Высшей школы экономики 2026 Т. 30 № 1 С. 128–153
This paper conducts comprehensive stress-testing of the sensitivity of a heuristic switching model. This model explains the phenomenon of fat tails in the distribution of macroeconomic data, which classic New Keynesian models fail to replicate. The solution lies in relaxing the assumption of full rationality. In the model, agents rely on a selection of simple ...
Added: April 7, 2026
Запорожец Д. Н., Simarova E., Записки научных семинаров ПОМИ РАН 2025 Т. 544 С. 130–153
Работа посвящена изучению асимптотических свойств случайных многогранников, порожденных выпуклыми оболочками независимых одинаково распределенных случайных векторов с правильно меняющимся распределением (с тяжелым хвостом). Исследуется сходимость функционалов данных случайных многогранников, включая внутренние объемы, порожденные ими U-max статистики и f-вектор, к соответствующим функционалам пуассоновских многогранников. Полученные результаты обобщают известные факты для отдельных распределений на общий класс распределений с ...
Added: February 2, 2026
Borodich E., Gasnikov A., Kovalev D., , in: Volume 267: International Conference on Machine Learning, 13-19 July 2025, Vancouver Convention Center, Vancouver, CanadaVol. 267.: [б.и.], 2025. P. 5045–5100.
Added: November 18, 2025
Gladin E., Alkousa M., Gasnikov A., Automation and Remote Control 2021 Vol. 82 P. 1679–1691
The article deals with some approaches to solving convex problems of the min-min type with smoothness and strong convexity in only one of the two groups of variables. It is shown that the proposed approaches based on Vaidya’s method, the fast gradient method, and the accelerated gradient method with variance reduction have linear convergence. It ...
Added: November 29, 2024
Gladin E., Gasnikov A., Ermakova E., Mathematical notes 2022 Vol. 112 No. 1 P. 183–190
The paper deals with a general problem of convex stochastic optimization in a space of small dimension (for example, 100 variables). It is known that for deterministic problems of convex optimization in small dimensions, the methods of centers of gravity type (for example, Vaidya’s method) provide the best convergence. For stochastic optimization problems, the question ...
Added: November 29, 2024
Gladin E., Gasnikov A., Dvurechensky P., Journal of Optimization Theory and Applications 2025 Vol. 204 No. 1 Article 1
Accuracy certificates for convex minimization problems allow for online verification of the accuracy of approximate solutions and provide a theoretically valid online stopping criterion. When solving the Lagrange dual problem, accuracy certificates produce a simple way to recover an approximate primal solution and estimate its accuracy. In this paper, we generalize accuracy certificates for the ...
Added: November 29, 2024
Gladin E., Зайнуллина К. Э., Компьютерные исследования и моделирование 2021 Т. 13 № 6 С. 1137–1147
The article considers minimization of the expectation of convex function. Problems of this type often arise in machine learning and a variety of other applications. In practice, stochastic gradient descent (SGD) and similar procedures are usually used to solve such problems. We propose to use the ellipsoid method with mini-batching, which converges linearly and can ...
Added: November 29, 2024
Rudenko V., Yudin N., Васин А. А., Компьютерные исследования и моделирование 2023 Т. 15 № 2 С. 329–353
This article reviews both historical achievements and modern results in the field of Markov Decision Process (MDP) and convex optimization. This review is the first attempt to cover the field of reinforcement learning in Russian in the context of convex optimization. The fundamental Bellman equation and the criteria of optimality of policy — strategies based on it, ...
Added: November 29, 2024
Kornilov N., Shamir O., Lobanov A. et al., , in: Advances in Neural Information Processing Systems 36 (NeurIPS 2023).: Curran Associates, Inc., 2023. P. 64083–64102.
Added: March 26, 2024
Beznosikov A., Richtarik P., Diskin M. et al., , in: Thirty-Sixth Conference on Neural Information Processing Systems : NeurIPS 2022.: Curran Associates, Inc., 2022. P. 14013–14029.
Added: January 27, 2023
Guminov S., Dvurechensky P., Tupitsa N. et al., , in: Proceedings of the 38th International Conference on Machine Learning (ICML 2021)Vol. 139.: PMLR, 2021. P. 3886–3898.
Added: October 30, 2022
Ivanova A., Dvurechensky P., Vorontsova E. et al., Journal of Optimization Theory and Applications 2022 Vol. 193 No. 1-3 P. 462–490
Many convex optimization problems have structured objective functions written as a sum of functions with different oracle types (e.g., full gradient, coordinate derivative, stochastic gradient) and different arithmetic operations complexity of these oracles. In the strongly convex case, these functions also have different condition numbers that eventually define the iteration complexity of first-order methods and ...
Added: October 28, 2022