Kuruzov Ilya Alekseevich

Born on January 18, 1999

Education

He completed his bachelor's degree at the Moscow Institute of Physics and Technology in 2020, his master's degree there in 2022, and defended his PhD thesis in 2025.

Work Experience

Since 2019, I have been a researcher at the Laboratory of Numerical Methods of Applied and Structural Optimization, and also worked at Laboratory No. 7 of the IPPI RAS. Since 2024, I have been working as a researcher at Innopolis University.

I am the author of several scientific articles on optimization methods with an inaccurate gradient and in a decentralized formulation. During his career, he participated in various commercial projects in the field of neural network training and signal processing.

In addition, I take an active part in the development of a commercial solver for integer programming problems: I optimize algorithms, increase the reliability and performance of the solver.

Teaching

Courses on optimization methods at MIPT from 2021, as well as at AI Masters from 2023

Research Interests

Inaccurate gradient optimization, decentralized optimization, adaptive optimization

Публикации

2025

Xiaokai Chen, Ilya Kuruzov, Gesualdo Scutari, Alexander Gasnikov. A parameter-free decentralized algorithm for composite convex optimization (2025) 2025 IEEE 64th Conference on Decision and Control (CDC) A

PDF

Ilya Kuruzov, Mohammad Alkousa, Fedor Stonyakin, Alexander Gasnikov. Gradient-type methods for decentralized optimization problems with Polyak–Łojasiewicz condition over time-varying networks (2025) Optimization Methods and Software, 1–28. DOI Q2

PDF

Ilya Kuruzov, Alexander Rogozin, Demyan Yarmoshik, Alexander Gasnikov. The mirror-prox sliding method for non-smooth decentralized saddle-point problems (2025) Optimization Methods and Software. DOI Q2

PDF

Ilya Kuruzov, Xiaokai Chen, Gesualdo Scutari, Alexander Gasnikov. Adaptive stepsize selection in decentralized convex optimization (2025) arXiv preprint arXiv:2507.23725

2024

Ilya Kuruzov, Gesualdo Scutari, Alexander Gasnikov. Achieving linear convergence with parameter-free algorithms in decentralized optimization (2024) Advances in Neural Information Processing Systems A*

PDF

Dorn, Y., Kornilov, N., Kutuzov, N., Nazin, A., Gorbunov, E., Gasnikov, A. Implicitly normalized forecaster with clipping for linear and non-linear heavy-tailed multi-armed bandits (2024) Computational Management Science, 21 (1), статья № 19. Scopus DOI Q3

PDF

Gasnikov, A.V., Alkousa, M.S., Lobanov, A.V., Dorn, Y.V., Stonyakin, F.S., Kuruzov, I.A., Singh, S.R. On Quasi-Convex Smooth Optimization Problems by a Comparison Oracle (2024) Russian Journal of Nonlinear Dynamics, 20 (5), pp. 813-825. Scopus DOI Q3

PDF

2023

Stonyakin F., Kuruzov I., Polyak B. Stopping Rules for Gradient Methods for Non-convex Problems with Additive Noise in Gradient (2023) Journal of Optimization Theory and Applications, 198 (2), pp. 531 - 551 Scopus DOI Q1

PDF

Alkousa, M.S., Gasnikov, A.V., Gladin, E.L., Kuruzov, I.A., Pasechnyuk, D.A., Stonyakin, F.S. Solving strongly convex-concave composite saddle-point problems with low dimension of one group of variables (2023) Sbornik Mathematics, 214 (3), pp. 285-333. Scopus DOI Q2

PDF