Publications

Slavomír Hanzely (2023). Sketch-and-Project Meets Newton Method: Global O(1/k^2) Convergence with Low-Rank Updates. arXiv:2305.13082.

PDF Cite

Konstantin Mishchenko, Slavomı́r Hanzely, Peter Richtárik (2023). Convergence of First-Order Algorithms for Meta-Learning with Moreau Envelopes. arXiv:2301.06806.

PDF Cite

Rustem Islamov, Xun Qian, Slavomír Hanzely, Mher Safaryan, Peter Richtárik (2022). Distributed Newton-Type Methods with Communication Compression and Bernoulli Aggregation. arxiv:2206.03588.

PDF Cite

Slavomír Hanzely, Dmitry Kamzolov, Dimitry Pasechnyuk, Alexander Gasnikov, Peter Richtárik, Martin Takáč (2022). A Damped Newton Method Achieves Global O(1/k^2) and Local Quadratic Convergence Rate. NeurIPS 2022.

PDF Cite

Zhize Li, Slavomír Hanzely, Peter Richtárik (2021). ZeroSARAH: Efficient Nonconvex Finite-Sum Optimization with Zero Full Gradient Computation. arxiv:2103.01447.

PDF Cite

Filip Hanzely, Slavomír Hanzely, Samuel Horváth, Peter Richtárik (2020). Lower Bounds and Optimal Algorithms for Personalized Federated Learning. NeurIPS 2020.

PDF Cite

Motasem Alfarra, Slavomír Hanzely, Alyazeed Albasyoni, Bernard Ghanem, Peter Richtárik (2020). Adaptive Learning of the Optimal Mini-Batch Size of SGD. OPT ML workshop, NeurIPS 2020.

PDF Cite