About

Publications

Contact

---

Publications

• Ilyas Fatkhullin, Niao He. Taming Nonconvex Stochastic Mirror Descent with General Bregman Divergence, AISTATS 2024.

• Ilyas Fatkhullin, Niao He, Yifan Hu. Stochastic Optimization under Hidden Convexity, OptML workshop at NeurIPS 2023.

• Jiduan Wu, Anas Barakat, Ilyas Fatkhullin, Niao He. Learning Zero-Sum Linear Quadratic Games with Improved Sample Complexity, Conference on Decision and Control 2023.

• Ilyas Fatkhullin, Alexander Tyurin, Peter Richtárik. Momentum Provably Improves Error Feedback!, NeurIPS 2023.

• Junchi Yang, Xiang Li, Ilyas Fatkhullin, Niao He. Two Sides of One Coin: the Limits of Untuned SGD and the Power of Adaptive Methods, NeurIPS 2023.

• Anas Barakat, Ilyas Fatkhullin, Niao He. Reinforcement Learning with General Utilities: Simpler Variance Reduction and Large State-Action Space, ICML 2023.

• Ilyas Fatkhullin, Anas Barakat, Anastasia Kireeva, Niao He. Stochastic Policy Gradient Methods: Improved Sample Complexity for Fisher-non-degenerate Policies, ICML 2023.

• Ilyas Fatkhullin, Jalal Etesami, Niao He, Negar Kiyavash. Sharp Analysis of Stochastic Optimization under Global Kurdyka-Łojasiewicz Inequality, NeurIPS 2022.

• Peter Richtárik, Igor Sokolov, Ilyas Fatkhullin, Elnur Gasanov, Eduard Gorbunov, Zhize Li. 3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation, ICML 2022 (spotlight).

• Ilyas Fatkhullin, Igor Sokolov, Eduard Gorbunov, Zhize Li, Peter Richtárik. EF21 with Bells & Whistles: Practical Algorithmic Extensions of Modern Error Feedback, OptML workshop at NeurIPS 2021.

• Peter Richtárik, Igor Sokolov, Ilyas Fatkhullin. EF21: A New, Simpler, Theoretically Better, and Practically Faster Error Feedback, NeurIPS 2021 (oral).

• Ilyas Fatkhullin, Boris Polyak. Optimizing Static Linear Feedback: Gradient Method, SIAM Journal on Control and Optimization 59, 3887-3911 (2021).

• Boris Polyak, Ilyas Fatkhullin. Use of Projective Coordinate Descent in the Fekete Problem, Comput. Math. and Math. Phys. 60, 795–807 (2020).