Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
Cooperative Multi-Agent Learning: The State of the Art
Autonomous Agents and Multi-Agent Systems
Reward shaping for valuing communications during multi-agent coordination
Proceedings of The 8th International Conference on Autonomous Agents and Multiagent Systems - Volume 1
Evolving large scale UAV communication system
Proceedings of the 14th annual conference on Genetic and evolutionary computation
| Hi-index | 0.00 |
Nash equilibrium (NE) assumes that players always make a best response. However, this is not always true; sometimes people cooperate even it is not a best response to do so. For example, in the Prisoner's Dilemma, people often cooperate. We consider two solution concepts that were introduced recently that try to capture such cooperation in two-player games: perfect cooperative equilibrium (PCE) (and an extension called maximum PCE (M-PCE)) [8] and the coco value [11]. We show that, despite their definitions being quite different, these notions are closely related, both in terms of axiomatization and algebraic characterization. We also consider the problem of computing how well players do when they cooperate according to these solution concepts, and show that in both cases in polynomial time. In the case of the coco value, this follows easily from the definition; in the case of the corresponding M-PCE value, it follows from a theorem showing that bilinear programming for a class of 2 X 2 matrices is in constant time, a result that may be of independent interest.