Introduction to Federated Optimization
Talk, Samsung AI Reading Club, Staines-upon-Thames, UK
Talks and presentations
Byzantine Robustness and Partial Participation Can Be Achieved Simultaneously: Just Clip Gradient Differences
Talk, Federated Learning One World Seminar (FLOW), Online
Byzantine Robustness and Partial Participation Can Be Achieved Simultaneously: Just Clip Gradient Differences
Talk, Rising Stars in AI Symposium 2024 at KAUST, Thuwal, Saudi Arabia
Server-side stepsizes and sampling without replacement provably help in federated optimization
Talk, 4th International Workshop on Distributed Machine Learning, Paris, France
Random reshuffling with variance reduction: New analysis and better rates
Poster, 39th Conference on Uncertainty in Artificial Intelligence, Online
Can 5th Generation Local Training Methods Support Client Sampling? Yes!
Talk, Third International Conference Mathematics in Armenia: Advances and Perspectives, Yerevan, Armenia
ProxSkip and its Variations: 5th Generation of Local Training Methods in Federated Learning
Talk, Google Research Seminar Zachary Charles), Online
Google Research Seminar (invited by Zachary Charles)
Can 5th Generation Local Training Methods Support Client Sampling? Yes!
Poster, 26th International Conference on Artificial Intelligence and Statistics (AISTATS), Valencia, Spain
Can 5th Generation Local Training Methods Support Client Sampling? Yes!
Poster, Rising Stars in AI Symposium 2023 at KAUST, Thuwal, Saudi Arabia
On 5th Generation of Local Training Methods in Federated Learning
Talk, MIPT Intelligent Systems Seminar, Online
ProxSkip: Breaking the Communication Complexity Barrier of Local Gradient Methods
Talk, 15th Viennese Conference on Optimal Control and Dynamic Games 2022, Vienna, Austria
ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!
Talk, EPFL Machine Learning and Optimization Laboratory Seminar, Lausanne, Switzerland
ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!
Talk, CISPA Helmholtz Center for Information Security seminar, Saarbrücken, Germany
ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!
Talk, WIAS Stochastic Algorithms and Nonparametric Statistics group seminar, online
ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!
Talk, UCL CORE (Yurii Nesterov's group) Optimization Seminar, Louvain-la-Neuve, Belgium
ProxSkip: Breaking the Communication Complexity Barrier of Local Gradient Methods
Talk, All-Russian Optimization Seminar, online
Server-Side Stepsizes and Sampling Without Replacement Provably Help in Federated Optimization
Talk, Federated Learning One World Seminar (FLOW), online
ProxSkip: Breaking the Communication Complexity Barrier of Local Gradient Methods
Talk, Rising Stars in AI Symposium 2022 at KAUST, Thuwal, KAUST
On Server-Side Stepsizes in Federated Optimization: Theory Explaining the Heuristics
Poster, International OPT Workshop on Optimization for Machine Learning, NeurIPS 2021, online
Better Linear Rates for SGD with Data Shuffling
Poster and Talk, International OPT Workshop on Optimization for Machine Learning, NeurIPS 2021, online
Federated Random Reshuffling with Compression and Variance Reduction
Poster, International Workshop on Federated Learning for User Privacy and Data Confidentiality, ICML 2021, online
Random Reshuffling with Variance Reduction New Analysis and Better Rates
Poster and Talk, Conference Optimization Without Borders, Voronovo, Russia
Random Reshuffling with Variance Reduction New Analysis and Better Rates
Talk, Traditional Youth School Control, Information and Optimization, Voronovo, Russia
Random Reshuffling with Variance Reduction New Analysis and Better Rates
Talk, KAUST Conference on Artificial Intelligence 2021, Thuwal, Saudi Arabia
Determination of data complexity using a universal approximating model
Talk, Mathematical Methods for Pattern Recognition: the 19th Russian National Conference with International Participation, Moscow, Russia
Book of abstracts (page 68).
Averaged Heavy Ball Method
Talk, 62th Scientific conference at MIPT, Section of Data analysis, recognition and prediction, Moscow, Russia
Best talk Award.
Averaged Heavy Ball Method
Poster, Traditional Youth School: Control, Information and Optimization, Voronovo, Russia
Best poster Award.