Examples of stanfordlogic.jocular.game.rlgamer.SkeletonEnvironment

Package stanfordlogic.jocular.game.rlgamer

Examples of stanfordlogic.jocular.game.rlgamer.SkeletonEnvironment

stanfordlogic.jocular.game.rlgamer.SkeletonEnvironment
This is a very simple environment with discrete observations corresponding to states labeled {0,1,...,19,20} The starting state is 10. There are 2 actions = {0,1}. 0 decrements the state, 1 increments the state. The problem is episodic, ending when state 0 or 20 is reached, giving reward -1 or +1, respectively. The reward is 0 on all other steps. @author Brian Tanner

    {
        super.setupInitialState();
        
        // Create the Agent
        theAgent = new SkeletonAgent();
        theEnvironment = new SkeletonEnvironment();
        RLGlue.setGlue(new LocalGlue(theEnvironment,theAgent));


        /* We could also start over and do another experiment */
        RLGlue.RL_init();

View Full Code Here

TOP

Related Classes of stanfordlogic.jocular.game.rlgamer.SkeletonEnvironment

org.rlcommunity.rlglue.codec.types.Observation

org.rlcommunity.rlglue.codec.types.Reward_observation_terminal

org.rlcommunity.rlglue.codec.util.EnvironmentLoader

stanfordlogic.jocular.game.RLGamer

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.