EAQR: A Multiagent Q-Learning Algorithm for Coordination of Multiple Agents

Table 8

Maximal steps for the DSN problem (evaluation episodes = 5000).

	= 10,000	= 50,000	= 100,000
EAQR	4.81	3.93	3.77
WoLF-PHC	5.34	5.34	5.67
EMA Q-learning	4.72	4.94	4.80
Single-agent RL	6.00	6.00	5.66