burlap.behavior.singleagent.learning.actorcritic

Class Actor

java.lang.Object

burlap.behavior.singleagent.learning.actorcritic.Actor

All Implemented Interfaces:

Policy

Direct Known Subclasses:

BoltzmannActor

public abstract class Actor
extends java.lang.Object
implements Policy

This class provides interface necessary for the actor portion of an Actor-Critic learning algorithm. Actors are almost entirely identical to policies since they effectively specify how the agent should act; in fact, this abstract class extends the Policy class. However, the extra important functionality that an actor must incorporate is the ability to adjust its policy in response to some critique of its behavior. In this class, this functionality should be implemented in the updateFromCritique(CritiqueResult) method.

Author:

James MacGlashan

Constructor Summary

Constructors

Constructor and Description
Actor()

Method Summary

Modifier and Type	Method and Description
abstract void	addNonDomainReferencedAction(ActionType a) This method allows the actor to utilize actions that are not apart of the domain definition.
abstract void	resetData() Used to reset any data that was created/modified during learning so that learning can be begin anew.
abstract void	updateFromCritique(CritiqueResult critqiue) Causes this object to update its behavior is response to a critique of its behavior.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface burlap.behavior.policy.Policy

action, actionProb, definedFor

Constructor Detail

Actor

public Actor()

Method Detail

updateFromCritique

public abstract void updateFromCritique(CritiqueResult critqiue)

Causes this object to update its behavior is response to a critique of its behavior.

Parameters:

critqiue - the critique of the agents behavior represented by a CritiqueResult object

addNonDomainReferencedAction

public abstract void addNonDomainReferencedAction(ActionType a)

This method allows the actor to utilize actions that are not apart of the domain definition.

Parameters:

a - an action not apart of the of the domain definition that this actor should be able to use.

resetData

public abstract void resetData()

Used to reset any data that was created/modified during learning so that learning can be begin anew.