Waltman, Ludo (contributor); Kaymak, Uzay (contributor) - 2006
environment. The
use of Q-learning is most appropriate if the environment is a Markov decision process. For such
an environment it … agents. An individual agent in multi-agent Q-learning perceives its environment as non-
stationary because the behavior of … the other agents changes over time due to learning. Since a
non-stationary environment does not have the properties of a …