Class i_StepLearner_id


public class i_StepLearner_id
extends NodeInt

A node that uses a reinforcement learning module to learn over time which output to select, given the current state and reward. Differs from i_Learner_id because it calls the learner each step.

For detailed information on how to configure behaviors, see the Clay page.

Copyright (c)1997, 1998 Tucker Balch

Fields inherited from class
DEBUG, embedded_nodes
Constructor Summary
i_StepLearner_id(i_ReinforcementLearner_id m, long ms, NodeInt e1, NodeScalar e2)
          Instantiate a reinforcement learning node with a specified learning module.
Method Summary
 int Value(long timestamp)
          Get the value of the node.
Methods inherited from class
booleanValue, doubleValue, intValue
Methods inherited from class
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail


public i_StepLearner_id(i_ReinforcementLearner_id m,
                        long ms,
                        NodeInt e1,
                        NodeScalar e2)
Instantiate a reinforcement learning node with a specified learning module.
m - i_ReinforcementLearning_id, the learning module.
ms - long, maximum time to stay in one state.
e1 - NodeInt, a node that provides state information.
e2 - NodeScalar, a node that provides the reinforcement signal.
Method Detail


public int Value(long timestamp)
Get the value of the node.
Value in class NodeInt
timestamp - long, the time of the request.