New answers tagged

1

Given that your state space is continuous, then I would recommend using Deep Q-Learning. As you say, running several episodes will definitely be beneficial so that the agent is able to explore the space more thoroughly.


Top 50 recent answers are included