Zobrazeno 1 - 10
of 27
pro vyhledávání: '"Khirirat, Sarit"'
We provide the first proof of convergence for normalized error feedback algorithms across a wide range of machine learning problems. Despite their popularity and efficiency in training deep neural networks, traditional analyses of error feedback algo
Externí odkaz:
http://arxiv.org/abs/2410.16871
Reinforcement learning has recently gained unprecedented popularity, yet it still grapples with sample inefficiency. Addressing this challenge, federated reinforcement learning (FedRL) has emerged, wherein agents collaboratively learn a single policy
Externí odkaz:
http://arxiv.org/abs/2404.10635
Data similarity assumptions have traditionally been relied upon to understand the convergence behaviors of federated learning methods. Unfortunately, this approach often demands fine-tuning step sizes based on the level of data similarity. When data
Externí odkaz:
http://arxiv.org/abs/2403.02347
Parallel stochastic gradient methods are gaining prominence in solving large-scale machine learning problems that involve data distributed across multiple nodes. However, obtaining unbiased stochastic gradients, which have been the focus of most theo
Externí odkaz:
http://arxiv.org/abs/2403.00853
Autor:
Khirirat, Sarit, Gorbunov, Eduard, Horváth, Samuel, Islamov, Rustem, Karray, Fakhri, Richtárik, Peter
Motivated by the increasing popularity and importance of large-scale training under differential privacy (DP) constraints, we study distributed gradient methods with gradient clipping, i.e., clipping applied to the gradients computed from local infor
Externí odkaz:
http://arxiv.org/abs/2305.18929
Federated learning (FL) is a distributed machine learning (ML) framework where multiple clients collaborate to train a model without exposing their private data. FL involves cycles of local computations and bi-directional communications between the c
Externí odkaz:
http://arxiv.org/abs/2304.05127
Zeroth-order methods have become important tools for solving problems where we have access only to function evaluations. However, the zeroth-order methods only using gradient approximations are $n$ times slower than classical first-order methods for
Externí odkaz:
http://arxiv.org/abs/2202.04612
Autor:
Khirirat, Sarit
Technological developments in devices and storages have made large volumes of data collections more accessible than ever. This transformation leads to optimization problems with massive data in both volume and dimension. In response to this trend, th
Externí odkaz:
http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-263738
With the increasing scale of machine learning tasks, it has become essential to reduce the communication between computing nodes. Early work on gradient compression focused on the bottleneck between CPUs and GPUs, but communication-efficiency is now
Externí odkaz:
http://arxiv.org/abs/2003.06377
The emergence of big data has caused a dramatic shift in the operating regime for optimization algorithms. The performance bottleneck, which used to be computations, is now often communications. Several gradient compression techniques have been propo
Externí odkaz:
http://arxiv.org/abs/1909.10327