Performance Bounded Reinforcement Learning in Strategic Interactions

44 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Empirically Evaluating Multiagent Learning Algorithms (2014, methodology): “wer. Instead, like Q-learning, it merely assumes that it is able to observe its own reward. RVσ(t) (Banerjee and Peng, 2006) belongs to a second line of gradient algorithms that started with ReDVaLeR(Banerjee and Peng, 2004). This algorithm also uses an adaptive step size when following the payoff g”
AWESOME: A general multiagent learning algorithm that converges in self-play and learns a best response against stationary opponents (2003, methodology): “Another algorithm, ReDVaLeR, was proposed more recently ( Banerjee & Peng, 2004 ) (after the introduction of the AWESOME algorithm, described in this paper, at the International Conference on Machine Learning, 2003).”
Convergence, Targeted Optimality, and Safety in Multiagent Learning (2010, background): “Subsequent approaches extended the rationality and convergence criteria to arbitrary (multi-player, multi-action) repeated games [1, 5].”
A Survey of Learning in Multiagent Environments: Dealing with Non-Stationarity (2017, background): “Moreover, ReDVaLeR adds another characteristic, constant bounded expected regret at any time against any number of opponents (Banerjee and Peng, 2004).”
Multiagent learning in the presence of memory-bounded agents (2013, unknown): “Subsequent approaches extended the rationality and convergence criteria to arbitrary (multi-player, multi-action) repeated games [3,16].”