Novitsky Anton Gennadievich
Education
Магистр
Work Experience
ISP RAS, since 2021
Teaching
Кафедра МОУ
Research Interests
(De)centralized optimization, second-order methods
Research Projects and Grants
Decentralized Cubic Newton with Consensus Procedure, together with Rogozin, Agafonov, Sokolov, and others.
Публикации
Artem Agafonov, Anton Novitskii, Alexander Rogozin, Yury Sokolov, Dmitry Kamzolov, Alexander Dyakonov, Martin Takáč, Alexander Gasnikov. Decentralized Inexact Cubic Newton Method with Consensus Procedure (2026). IEEE Access, 14, 83513–83540. DOI: 10.1109/ACCESS.2026.3695635. Q1
Anton Novitskii, Alexander Rogozin, Alexander Gasnikov. Decentralized Gradient Descent with an Increasing Consensus Procedure Accuracy (2026). MOTOR 2026. Lecture Notes in Computer Science, 16687, 48–63. DOI: 10.1007/978-3-032-31390-4_4. Q2
Darina M. Dvinskikh, Anton G. Novitskii, Ewsey R. Obzherin, Alexander V. Gasnikov. TIGER: Tunable-Geometry Accelerated Gradient Sliding with Adaptive Safeguards under Similarity (2026). arXiv preprint arXiv:2610.25504.
Darina Dvinskikh, Anton Novitskii, Demyan Yarmoshik, Alexander Rogozin, Vladislav Matyukhin, Alexander Gasnikov. Accelerated Consensus and Decentralized Optimization over Slowly Varying Networks (2026). arXiv preprint arXiv:2609.33824.