UC Irvine & DeepMind’s Anytime Optimal PSRO: Guaranteed Convergence to a Nash Equilibrium With Decreased Exploitability in Two-Player Zero-Sum Games

A research team from the University of California Irvine and DeepMind proposes Anytime Optimal PSRO, a new PSRO variant for two-player zero-sum games that is guaranteed to converge to a Nash equilibrium while decreasing exploitability from iteration to iteration.

ai artificial intelligence deepmind equilibrium games machine learning machine learning & data science ml nash equilibrium reinforcement learning research technology zero-sum game