Pronađeno: 1-3 / 3 radova

Autori: Yu Shuhua

>> Filter: Samo Article i Review

>> Sve godine

Naslov Smoothed Gradient Clipping and Error Feedback for Decentralized Optimization Under Symmetric Heavy-Tailed Noise (Article)
Autori Yu Shuhua Jakovetic Dusan Kar Soummya 
Info SIAM JOURNAL ON OPTIMIZATION, (2026), vol. 36 br. 2, str. 703-728
Projekat NSF [ECCS 2330196]; Ministry of Science, Technological Development and Innovation, Republic of Serbia; Science Fund of Republic of Serbia [7359]; Provincial Secretariat for Higher Education and Scientific Research [142-451-2593/2021-01/2]
Ispravka Web of Science   Članak   Elečas   Rang časopisa  
facebook sharing button
twitter sharing button
linkedin sharing button
gmail sharing button
copy sharing button
Naslov Large Deviation Upper Bounds and Improved Mse Rates of Nonlinear Sgd: Heavy-Tailed Noise and Power of Symmetry (Article)
Autori Armacki Aleksandar Yu Shuhua Bajovic Dragana  Jakovetic Dusan Kar Soummya 
Info SIAM JOURNAL ON OPTIMIZATION, (2026), vol. 36 br. 1, str. 32-59
Projekat National Science Foundation [ECCS 2330196]; Ministry of Science, Technological Development, and Innovation [451-03-65/2024-03/200156]; Faculty of Technical Sciences, University of Novi Sad [01-3394/1]; Ministry of Science, Technological Development, and Innovation of the Republic of Serbia [451-03-137/2025-03/200125, 451-03-136/2025-03/200125]; Science Fund of Republic of Serbia, project "LASCADO" [7359]; Provincial Secretariat for Higher Education and Scientific Research [142-451-2593/2021-01/2]
Ispravka Web of Science   Članak   Elečas   Rang časopisa  
facebook sharing button
twitter sharing button
linkedin sharing button
gmail sharing button
copy sharing button
Naslov High-probability Convergence Bounds for Online Nonlinear Stochastic Gradient Descent under Heavy-tailed Noise (Proceedings Paper)
Autori Armacki Aleksandar Yu Shuhua Sharma Pranay Joshi Gauri Bajovic Dragana  Jakovetic Dusan Kar Soummya 
Info INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, (2025), vol. 258 br. , str. -
Projekat NSF [CCF 2045694, CNS-2112471, CPS-2111751, SHF-2107024, ECCS-2330196]; ONR [N00014-23-1-2149]; Ministry of Science, Technological Development and Innovation [451-03-65/2024-03/200156]; Faculty of Technical Sciences, University of Novi Sad [01-3394/1]; European Union's Horizon Europe program [101093006, 101135775, 101135916]; Science Fund of Republic of Serbia, project "LASCADO" [7359]; Provincial Secretariat for Higher Education and Scientific Research [142-451-2593/2021-01/2]
Ispravka Web of Science  
facebook sharing button
twitter sharing button
linkedin sharing button
gmail sharing button
copy sharing button
>> Sve godine

Ispis zapisa u formatu:TXT | BibTeX