Deep Q-Network with Predictive State Models in Partially Observable Domains

Table 2

The best mean reward of three methods.

	CartPole-v1	Swimmer-v1	Reacher-v1
DRQN	189	30.23	−11.25
DQN-1frame	46.76	23.78	−17.78
RPSR-DQN	194	38.51	−9.24
RPSP	116	21.32	−70.23