Filip Hanzely
Filip Hanzely
Research Assistant Professor, TTIC
Bestätigte E-Mail-Adresse bei - Startseite
Zitiert von
Zitiert von
A unified theory of SGD: Variance reduction, sampling, quantization and coordinate descent
E Gorbunov, F Hanzely, P Richtárik
International Conference on Artificial Intelligence and Statistics, 680-690, 2020
Federated learning of a mixture of global and local models
F Hanzely, P Richtárik
arXiv preprint arXiv:2002.05516, 2020
Accelerated Bregman proximal gradient methods for relatively smooth convex optimization
F Hanzely, P Richtarik, L Xiao
arXiv preprint arXiv:1808.03045, 2018
Accelerated stochastic matrix inversion: general theory and speeding up BFGS rules for faster second-order optimization
RM Gower, F Hanzely, P Richtárik, S Stich
arXiv preprint arXiv:1802.04079, 2018
SEGA: Variance reduction via gradient sketching
F Hanzely, K Mishchenko, P Richtárik
arXiv preprint arXiv:1809.03054, 2018
Fastest rates for stochastic mirror descent methods
F Hanzely, P Richtárik
Computational Optimization and Applications, 1-50, 2021
Accelerated coordinate descent with arbitrary sampling and best rates for minibatches
F Hanzely, P Richtárik
AISTATS 2019, 2019
Testing for causality in reconstructed state spaces by an optimized mixed prediction method
A Krakovská, F Hanzely
Physical Review E 94 (5), 052203, 2016
99% of worker-master communication in distributed optimization is not needed
K Mishchenko, F Hanzely, P Richtárik
Conference on Uncertainty in Artificial Intelligence, 979-988, 2020
Privacy preserving randomized gossip algorithms
F Hanzely, J Konečný, N Loizou, P Richtárik, D Grishchenko
arXiv preprint arXiv:1706.07636, 2017
Lower bounds and optimal algorithms for personalized federated learning
F Hanzely, S Hanzely, S Horváth, P Richtárik
NeurIPS 2020, 2020
A nonconvex projection method for robust PCA
A Dutta, F Hanzely, P Richtárik
Proceedings of the AAAI Conference on Artificial Intelligence 33 (01), 1468-1476, 2019
One method to rule them all: Variance reduction for data, parameters and many new methods
F Hanzely, P Richtárik
arXiv preprint arXiv:1905.11266, 2019
Stochastic Subspace Cubic Newton Method
F Hanzely, N Doikov, P Richtárik, Y Nesterov
ICML 2020, 2020
Local sgd: Unified theory and new efficient methods
E Gorbunov, F Hanzely, P Richtárik
International Conference on Artificial Intelligence and Statistics, 3556-3564, 2021
A privacy preserving randomized gossip algorithm via controlled noise insertion
F Hanzely, J Konečný, N Loizou, P Richtárik, D Grishchenko
NeurIPS PPML workshop 2018, 2018
Best pair formulation & accelerated scheme for non-convex principal component pursuit
A Dutta, F Hanzely, J Liang, P Richtárik
IEEE Transactions on Signal Processing 68, 6128-6141, 2020
Variance Reduced Coordinate Descent with Acceleration: New Method With a Surprising Application to Finite-Sum Problems
F Hanzely, D Kovalev, P Richtarik
ICML 2020, 2020
Personalized federated learning: A unified framework and universal optimization techniques
F Hanzely, B Zhao, M Kolar
arXiv preprint arXiv:2102.09743, 2021
Smoothness Matrices Beat Smoothness Constants: Better Communication Compression Techniques for Distributed Optimization
M Safaryan, F Hanzely, P Richtárik
arXiv preprint arXiv:2102.07245, 2021
Das System kann den Vorgang jetzt nicht ausführen. Versuchen Sie es später erneut.
Artikel 1–20