Symmetric (optimistic) natural policy gradient for multi-agent learning with parameter convergence

S Pattathil, K Zhang, A Ozdaglar - … Conference on Artificial …, 2023 - proceedings.mlr.press
Multi-agent interactions are increasingly important in the context of reinforcement learning,
and the theoretical foundations of policy gradient methods have attracted surging research …

Performative prediction with neural networks

M Mofakhami, I Mitliagkas… - … Conference on Artificial …, 2023 - proceedings.mlr.press
Performative prediction is a framework for learning models that influence the data they
intend to predict. We focus on finding classifiers that are performatively stable, ie optimal for …

Learning Equilibria in Adversarial Team Markov Games: A Nonconvex-Hidden-Concave Min-Max Optimization Problem

F Kalogiannis, J Yan, I Panageas - arXiv preprint arXiv:2410.05673, 2024 - arxiv.org
We study the problem of learning a Nash equilibrium (NE) in Markov games which is a
cornerstone in multi-agent reinforcement learning (MARL). In particular, we focus on infinite …

Stochastic optimization under hidden convexity

I Fatkhullin, N He, Y Hu - arXiv preprint arXiv:2401.00108, 2023 - arxiv.org
In this work, we consider constrained stochastic optimization problems under hidden
convexity, ie, those that admit a convex reformulation via non-linear (but invertible) map $ c …

Exploiting hidden structures in non-convex games for convergence to Nash equilibrium

I Sakos, EV Vlatakis-Gkaragkounis… - Advances in …, 2024 - proceedings.neurips.cc
A wide array of modern machine learning applications–from adversarial models to multi-
agent reinforcement learning–can be formulated as non-cooperative games whose Nash …

Solving hidden monotone variational inequalities with surrogate losses

R D'Orazio, D Vucetic, Z Liu, JL Kim, I Mitliagkas… - arXiv preprint arXiv …, 2024 - arxiv.org
Deep learning has proven to be effective in a wide variety of loss minimization problems.
However, many applications of interest, like minimizing projected Bellman error and min …

Position: Optimization in SciML Should Employ the Function Space Geometry

J Müller, M Zeinhofer - Forty-first International Conference on Machine … - openreview.net
We provide an infinite-dimensional view on optimization problems encountered in scientific
machine learning (SciML) and advocate for the paradigm first optimize, then discretize for …

Optimization and Generalization of Minimax Algorithms

S Pattathil - 2023 - dspace.mit.edu
This thesis explores minimax formulations of machine learning and multi-agent learning
problems, focusing on algorithmic optimization and generalization performance. The first …