Zobrazeno 1 - 10
of 3 060
pro vyhledávání: '"A A, Gasanov"'
Error Feedback (EF) is a highly popular and immensely effective mechanism for fixing convergence issues which arise in distributed training methods (such as distributed GD or SGD) when these are enhanced with greedy communication compression techniqu
Externí odkaz:
http://arxiv.org/abs/2402.10774
Autor:
Koldasbayeva, Diana, Tregubova, Polina, Gasanov, Mikhail, Zaytsev, Alexey, Petrovskaia, Anna, Burnaev, Evgeny
With the rise of electronic data, particularly Earth observation data, data-based geospatial modelling using machine learning (ML) has gained popularity in environmental research. Accurate geospatial predictions are vital for domain research based on
Externí odkaz:
http://arxiv.org/abs/2311.11057
Autor:
U. D. Belchenko, E. E. Yakubova, T. S. Popovich, E. L. Agaeva, E. Z. Kayalieva, E.-S. Sh. Khudaiberdiev, Y. R. Li, S. R. Gasanov, N. I. Perederenko, E. Sh. Kadyrov, K. V. Novikov, E. A. Ostozhiev, N. S. Prasolov, L. E. Sorokina
Publikováno v:
Акушерство, гинекология и репродукция, Vol 18, Iss 4, Pp 464-474 (2024)
Aim: systemically assessed characteristics of intestinal microbiota taxonomic composition in relation to parameters of hormonal and immune status in patients with external genital endometriosis (EGE).Materials and Methods. The controlled cross-sectio
Externí odkaz:
https://doaj.org/article/74a18da21fc944229d2a975522fd9503
Autor:
Daria A. Emelina, Ilya V. Kravchenko, Igor V. Makarov, Rauf F. Gasanov, Ekaterina S. Prokhorenko
Publikováno v:
Вопросы современной педиатрии, Vol 23, Iss 4, Pp 204-212 (2024)
There has been significant increase in mental disorders prevalence in pediatric population around the world. Increasing incidence of autism, intellectual incapacity, hyperkinetic disorders, and schizophrenia gives causes for specific concern. Clarify
Externí odkaz:
https://doaj.org/article/195c3918f4a94ae08faf0f6210743cc5
We propose a Randomized Progressive Training algorithm (RPT) -- a stochastic proxy for the well-known Progressive Training method (PT) (Karras et al., 2017). Originally designed to train GANs (Goodfellow et al., 2014), PT was proposed as a heuristic,
Externí odkaz:
http://arxiv.org/abs/2306.03626
We provide the first proof that gradient descent $\left({\color{green}\sf GD}\right)$ with greedy sparsification $\left({\color{green}\sf TopK}\right)$ and error feedback $\left({\color{green}\sf EF}\right)$ can obtain better communication complexity
Externí odkaz:
http://arxiv.org/abs/2305.15264
We propose Adaptive Compressed Gradient Descent (AdaCGD) - a novel optimization algorithm for communication-efficient training of supervised machine learning models with adaptive compression level. Our approach is inspired by the recently proposed th
Externí odkaz:
http://arxiv.org/abs/2211.00188
Autor:
D. A. Egorova, V. V. Derezina, M. V. Chebanyan, M. B. Sultonova, T. V. Ishmuratov, M. M. Gasanov, A. Z. Chipchikova, S. A. Paksileva, S. S. Mitkina, R. R. Khamidullina, A. Sh. Kutueva, A. A. Anikeeva
Publikováno v:
Акушерство, гинекология и репродукция, Vol 18, Iss 1, Pp 68-82 (2024)
Introduction. Today, infertility is a global problem that affects about 48.5 million married couples worldwide. It has been suggested that epigenetic aberrations are of great importance for reproductive health, as they account for an interactive rela
Externí odkaz:
https://doaj.org/article/05b5f31e98a34466b1cd1761f811af9e
Publikováno v:
Неотложная медицинская помощь, Vol 12, Iss 4, Pp 650-657 (2024)
Mechanical damage to the esophagus is a severe urgent pathology characterized by the development of life-threatening complications. With all the variety of modern diagnostic and treatment technologies, the mortality rate in this category of patients
Externí odkaz:
https://doaj.org/article/4aa5fe48dca14118ac07c6183a61ae92
Autor:
Richtárik, Peter, Sokolov, Igor, Fatkhullin, Ilyas, Gasanov, Elnur, Li, Zhize, Gorbunov, Eduard
We propose and study a new class of gradient communication mechanisms for communication-efficient training -- three point compressors (3PC) -- as well as efficient distributed nonconvex optimization algorithms that can take advantage of them. Unlike
Externí odkaz:
http://arxiv.org/abs/2202.00998