Learning equilibrium as a generalization of learning to optimize

摘要

We argue that learning equilibrium is an appropriate generalization to multi-agent systems of the concept of learning to optimize in single-agent setting. We further define and discuss the concept of weak learning equilibrium.