Ilyas Fatkhullin

31 papers A* 8A 2Journal 20Unranked 1

Year	Rank	Type	Title / Venue / Authors
2025	J	`jnl`	Can SGD Handle Heavy-Tailed Noise? CoRR Ilyas Fatkhullin, Florian Hübler, Guanghui Lan
2025	J	`jnl`	EF21 with Bells & Whistles: Six Algorithmic Extensions of Modern Error Feedback. J. Mach. Learn. Res. Ilyas Fatkhullin, Igor Sokolov, Eduard Gorbunov, Zhize Li, Peter Richtárik
2025	A	`conf`	From Gradient Clipping to Normalization for Heavy Tailed SGD. AISTATS Florian Hübler, Ilyas Fatkhullin, Niao He
2025	J	`jnl`	Global Solutions to Non-Convex Functional Constrained Problems with Hidden Convexity. CoRR Ilyas Fatkhullin, Niao He, Guanghui Lan, Florian Wolf
2025	J	`jnl`	Learning Zero-Sum Linear Quadratic Games with Improved Sample Complexity and Last-Iterate Convergence. SIAM J. Control. Optim. Jiduan Wu, Anas Barakat, Ilyas Fatkhullin, Niao He
2025	J	`jnl`	Natural Gradient VI: Guarantees for Non-Conjugate Models. CoRR Fangyuan Sun, Ilyas Fatkhullin, Niao He
2025	A*	`conf`	Safe-EF: Error Feedback for Non-smooth Constrained Optimization. ICML Rustem Islamov, Yarden As, Ilyas Fatkhullin
2025	J	`jnl`	Safe-EF: Error Feedback for Nonsmooth Constrained Optimization. CoRR Rustem Islamov, Yarden As, Ilyas Fatkhullin
2025	J	`jnl`	Second-order Optimization under Heavy-Tailed Noise: Hessian Clipping and Sample Complexity Limits. CoRR Abdurakhmon Sadiev, Peter Richtárik, Ilyas Fatkhullin
2025	J	`jnl`	Stochastic Optimization under Hidden Convexity. SIAM J. Optim. Ilyas Fatkhullin, Niao He, Yifan Hu
2024	J	`jnl`	From Gradient Clipping to Normalization for Heavy Tailed SGD. CoRR Florian Hübler, Ilyas Fatkhullin, Niao He
2024	J	`jnl`	Stochastic Optimization under Hidden Convexity. CoRR Ilyas Fatkhullin, Niao He, Yifan Hu
2024	A	`conf`	Taming Nonconvex Stochastic Mirror Descent with General Bregman Divergence. AISTATS Ilyas Fatkhullin, Niao He
2024	J	`jnl`	Taming Nonconvex Stochastic Mirror Descent with General Bregman Divergence. CoRR Ilyas Fatkhullin, Niao He
2023	—	`conf`	Learning Zero-Sum Linear Quadratic Games with Improved Sample Complexity. CDC Jiduan Wu, Anas Barakat, Ilyas Fatkhullin, Niao He
2023	J	`jnl`	Learning Zero-Sum Linear Quadratic Games with Improved Sample Complexity. CoRR Jiduan Wu, Anas Barakat, Ilyas Fatkhullin, Niao He
2023	A*	`conf`	Momentum Provably Improves Error Feedback! NeurIPS Ilyas Fatkhullin, Alexander Tyurin, Peter Richtárik
2023	J	`jnl`	Momentum Provably Improves Error Feedback! CoRR Ilyas Fatkhullin, Alexander Tyurin, Peter Richtárik
2023	A*	`conf`	Reinforcement Learning with General Utilities: Simpler Variance Reduction and Large State-Action Space. ICML Anas Barakat, Ilyas Fatkhullin, Niao He
2023	J	`jnl`	Reinforcement Learning with General Utilities: Simpler Variance Reduction and Large State-Action Space. CoRR Anas Barakat, Ilyas Fatkhullin, Niao He
2023	A*	`conf`	Stochastic Policy Gradient Methods: Improved Sample Complexity for Fisher-non-degenerate Policies. ICML Ilyas Fatkhullin, Anas Barakat, Anastasia Kireeva, Niao He
2023	J	`jnl`	Stochastic Policy Gradient Methods: Improved Sample Complexity for Fisher-non-degenerate Policies. CoRR Ilyas Fatkhullin, Anas Barakat, Anastasia Kireeva, Niao He
2023	A*	`conf`	Two Sides of One Coin: the Limits of Untuned SGD and the Power of Adaptive Methods. NeurIPS Junchi Yang, Xiang Li, Ilyas Fatkhullin, Niao He
2023	J	`jnl`	Two Sides of One Coin: the Limits of Untuned SGD and the Power of Adaptive Methods. CoRR Junchi Yang, Xiang Li, Ilyas Fatkhullin, Niao He
2022	A*	`conf`	3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation. ICML Peter Richtárik, Igor Sokolov, Elnur Gasanov, Ilyas Fatkhullin, Zhize Li, Eduard Gorbunov
2022	J	`jnl`	3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation. CoRR Peter Richtárik, Igor Sokolov, Ilyas Fatkhullin, Elnur Gasanov, Zhize Li, Eduard Gorbunov
2022	A*	`conf`	Sharp Analysis of Stochastic Optimization under Global Kurdyka-Lojasiewicz Inequality. NeurIPS Ilyas Fatkhullin, Jalal Etesami, Niao He, Negar Kiyavash
2021	J	`jnl`	EF21 with Bells & Whistles: Practical Algorithmic Extensions of Modern Error Feedback. CoRR Ilyas Fatkhullin, Igor Sokolov, Eduard Gorbunov, Zhize Li, Peter Richtárik
2021	A*	`conf`	EF21: A New, Simpler, Theoretically Better, and Practically Faster Error Feedback. NeurIPS Peter Richtárik, Igor Sokolov, Ilyas Fatkhullin
2021	J	`jnl`	EF21: A New, Simpler, Theoretically Better, and Practically Faster Error Feedback. CoRR Peter Richtárik, Igor Sokolov, Ilyas Fatkhullin
2021	J	`jnl`	Optimizing Static Linear Feedback: Gradient Method. SIAM J. Control. Optim. Ilyas Fatkhullin, Boris T. Polyak