Fast Last-Iterate Convergence of Learning in Games Requires Forgetful Algorithms.
Yang CaiGabriele FarinaJulien Grand-ClémentChristian KroerChung-Wei LeeHaipeng LuoWeiqiang ZhengPublished in: CoRR (2024)
Keyphrases
- learning algorithm
- learning process
- learning agents
- active learning
- noise tolerant
- reinforcement learning
- learning tasks
- learning systems
- theoretical analysis
- learning problems
- global convergence
- prior knowledge
- machine learning algorithms
- supervised learning
- stochastic games
- significant improvement
- stochastic approximation
- board game