Java Deep Learning Cookbook

How to do it...

1. Create an MDP wrapper to load the mission:

Sample: MalmoEnv mdp = new MalmoEnv("cliff_walking_rl4j.xml", actionSpace, observationSpace, obsPolicy);

2. Evaluate the agent:

Sample: double rewards = 0; for (int i = 0; i < 10; i++) { double reward = pol.play(mdp, new HistoryProcessor(MALMO_HPROC)); rewards += reward; Logger.getAnonymousLogger().info("Reward: " + reward); }