Asymmetric multiagent reinforcement learning
Web Intelligence and Agent Systems
Gradient descent for symmetric and asymmetric multiagent reinforcement learning
Web Intelligence and Agent Systems
Vector Valued Markov Decision Process for robot platooning
Proceedings of the 2008 conference on ECAI 2008: 18th European Conference on Artificial Intelligence
From cognition to docition: The teaching radio paradigm for distributed & autonomous deployments
Computer Communications
Hi-index | 0.00 |
A novel method for asymmetric multiagent reinforcement learning isintroduced in this paper. The method addresses the problem wherethe information states of the agents involved in the learning taskare not equal; some agents (leaders)have information how theiropponents (followers) will select their actions and based on thisinformation leaders encourage followers to select actions that leadto improved payoffs for the leaders. This kind of configurationarises e.g. in semi-centralized multiagent systems with an externalglobal utility associated to the system. We present a briefliterature survey of multi agent reinforcement learning based onMarkov games and then construct an asymmetric learning method thatutilizes the theory of Markov games. Additionally, we test theproposed method with a simple example application.