Следене
Ioannis Panageas
Ioannis Panageas
Assistant Professor, University of California, Irvine
Потвърден имейл адрес: ics.uci.edu - Начална страница
Заглавие
Позовавания
Позовавания
Година
First-order methods almost always avoid strict saddle points
JD Lee, I Panageas, G Piliouras, M Simchowitz, MI Jordan, B Recht
Mathematical programming 176 (1-2), 311-337, 2019
3782019
The limit points of (optimistic) gradient descent in min-max optimization
C Daskalakis, I Panageas
Advances in Neural Information Processing Systems 31, 9236-9246, 2018
3072018
Gradient descent only converges to minimizers: Non-isolated critical points and invariant regions
I Panageas, G Piliouras
Proceedings of the Conference on Innovations in Theoretical Computer Science, 2017
198*2017
Last-iterate convergence: Zero-sum games and constrained min-max optimization
C Daskalakis, I Panageas
Proceedings of the Conference on Innovations in Theoretical Computer Science, 2019
1852019
Multiplicative Weights Update with Constant Step-Size in Congestion Games: Convergence, Limit Cycles and Chaos
G Palaiopanos, I Panageas, G Piliouras
31st Annual Conference on Neural Information Processing Systems, 2017
1412017
Global convergence of multi-agent policy gradient in markov potential games
S Leonardos, W Overman, I Panageas, G Piliouras
arXiv preprint arXiv:2106.01969, 2021
1402021
First-order methods almost always avoid saddle points: The case of vanishing step-sizes
I Panageas, G Piliouras, X Wang
Advances in Neural Information Processing Systems 32, 2019
642019
Natural selection as an inhibitor of genetic diversity: Multiplicative weights updates algorithm and a conjecture of haploid genetics [working paper abstract]
R Mehta, I Panageas, G Piliouras
Proceedings of the 2015 Conference on Innovations in Theoretical Computer …, 2015
622015
Independent natural policy gradient always converges in markov potential games
R Fox, SM Mcaleer, W Overman, I Panageas
International Conference on Artificial Intelligence and Statistics, 4414-4425, 2022
582022
Last iterate convergence in no-regret learning: constrained min-max optimization for convex-concave landscapes
Q Lei, SG Nagarajan, I Panageas, X Wang
International Conference on Artificial Intelligence and Statistics, 1441-1449, 2021
502021
Average case performance of replicator dynamics in potential games via computing regions of attraction
I Panageas, G Piliouras
Proceedings of the 2016 ACM Conference on Economics and Computation, 703-720, 2016
50*2016
On last-iterate convergence beyond zero-sum games
I Anagnostides, I Panageas, G Farina, T Sandholm
International Conference on Machine Learning, 536-581, 2022
432022
Regression from dependent observations
C Daskalakis, N Dikkala, I Panageas
Proceedings of the 51st Annual ACM symposium on Theory of Computing, 2019
402019
Depth-width trade-offs for relu networks via sharkovsky's theorem
V Chatziafratis, SG Nagarajan, I Panageas, X Wang
arXiv preprint arXiv:1912.04378, 2019
312019
Fast convergence of langevin dynamics on manifold: Geodesics meet log-sobolev
X Wang, Q Lei, I Panageas
Advances in Neural Information Processing Systems 33, 18894-18904, 2020
232020
Cycles in zero-sum differential games and biological diversity
T Mai, M Mihail, I Panageas, W Ratcliff, V Vazirani, P Yunker
Proceedings of the 2018 ACM Conference on Economics and Computation, 339-350, 2018
23*2018
On the Analysis of EM for truncated mixtures of two Gaussians
SG Nagarajan, I Panageas
Algorithmic Learning Theory, 634-659, 2020
222020
On the convergence of no-regret learning dynamics in time-varying games
I Anagnostides, I Panageas, G Farina, T Sandholm
Advances in Neural Information Processing Systems 36, 2024
192024
Efficiently computing nash equilibria in adversarial team markov games
F Kalogiannis, I Anagnostides, I Panageas, EV Vlatakis-Gkaragkounis, ...
arXiv preprint arXiv:2208.02204, 2022
192022
Better depth-width trade-offs for neural networks through the lens of dynamical systems
V Chatziafratis, SG Nagarajan, I Panageas
International Conference on Machine Learning, 1469-1478, 2020
182020
Системата не може да изпълни операцията сега. Опитайте отново по-късно.
Статии 1–20