%0 Report %1 Kaisers2007 %A Kaisers, Michael %D 2007 %K FMQ-learning,Q-learning,iterated Q- games,learning,lenient games,reinforcement iterated learning learning,lenient q-,reinforcement %T Reinforcement Learning in Multi-agent Games - A value iteration perspective %X This article investigates the performance of independent reinforcement learners in multi- agent games. Convergence to Nash equilibria and parameter settings for desired learning be- havior are discussed for Q-learning, Frequency Maximum Q value (FMQ) learning and lenient Q-learning. FMQ and lenient Q-learning are shown to outperform regular Q-learning significantly in the context of coordination games with mis- coordination penalties. Furthermore, Q- learning with an $\epsilon$-greedy and FMQ learning with a Boltzmann action selection are shown to scale well to games with one thousand agents.

@techreport{Kaisers2007, abstract = {This article investigates the performance of independent reinforcement learners in multi- agent games. Convergence to Nash equilibria and parameter settings for desired learning be- havior are discussed for Q-learning, Frequency Maximum Q value (FMQ) learning and lenient Q-learning. FMQ and lenient Q-learning are shown to outperform regular Q-learning significantly in the context of coordination games with mis- coordination penalties. Furthermore, Q- learning with an $\epsilon$-greedy and FMQ learning with a Boltzmann action selection are shown to scale well to games with one thousand agents.}, added-at = {2016-12-19T12:09:05.000+0100}, author = {Kaisers, Michael}, biburl = {https://www.bibsonomy.org/bibtex/2b93e5e8ce8a0678f009b53f5616f302c/swarmlab}, interhash = {74a07bde6309691f725cd14921ec4df6}, intrahash = {b93e5e8ce8a0678f009b53f5616f302c}, keywords = {FMQ-learning,Q-learning,iterated Q- games,learning,lenient games,reinforcement iterated learning learning,lenient q-,reinforcement}, school = {Maastricht University}, timestamp = {2016-12-19T12:18:59.000+0100}, title = {{Reinforcement Learning in Multi-agent Games - A value iteration perspective}}, year = 2007 }