NTR organizes and hosts scientific webinars on neural networks and invites speakers from all over the world to present their recent work.
On February 1 Alexander Gasnikov, Moscow Institute of Physics and Technology, Moscow, Russia, presented a technical Zoom webinar on Saddle-point problems and efficient algorithms.
About the webinar:
During the last five years the great interest in saddle point problems is associated with the popularity of generative adversarial networks.
During the presentation, I reviewed some of the recent advances in the development of efficient numerical methods for solving such problems. I also described several open problems.
The main focus was specifically to issues related to theory (applications can be found in the articles cited during the presentation).
Related materials:
https://arxiv.org/pdf/2112.15199.pdf
https://proceedings.neurips.cc/paper/2021/file/44e65d3e9bc2f88b2b3d566de51a5381-Paper.pdf
https://arxiv.org/pdf/2110.03313.pdf
https://arxiv.org/pdf/2106.08315.pdf
https://arxiv.org/pdf/2106.07289.pdf
https://arxiv.org/pdf/2102.07758.pdf
https://arxiv.org/pdf/2010.13112.pdf
https://arxiv.org/pdf/2005.05913.pdf
Moderator and contact: NTR CEO Nick Mikhailovsky: nickm@ntrlab.com.