burlap.behavior.stochasticgames.agents.maql

Class MAQLFactory.MAMaxQLearningFactory

java.lang.Object

burlap.behavior.stochasticgames.agents.maql.MAQLFactory

burlap.behavior.stochasticgames.agents.maql.MAQLFactory.MAMaxQLearningFactory

All Implemented Interfaces:

AgentFactory

Enclosing class:

MAQLFactory

public static class MAQLFactory.MAMaxQLearningFactory
extends MAQLFactory

Factory for generating Max multiagent Q-learning agents. This is also known as "maxmax-Q" or "friend-Q." Fixes the backup operator to the max operator and the policy to an epsilon-greedy policy.

Author:

James MacGlashan

Nested Class Summary

Nested classes/interfaces inherited from class burlap.behavior.stochasticgames.agents.maql.MAQLFactory

MAQLFactory.CoCoQLearningFactory, MAQLFactory.MAMaxQLearningFactory

Field Summary

Fields inherited from class burlap.behavior.stochasticgames.agents.maql.MAQLFactory

backupOperator, discount, domain, hashingFactory, learningPolicy, learningRate, qInit, queryOtherAgentsQSource

Constructor Summary

Constructors

Constructor and Description
MAQLFactory.MAMaxQLearningFactory(SGDomain d, double discount, LearningRate learningRate, HashableStateFactory hashFactory, ValueFunctionInitialization qInit, boolean queryOtherAgentsForTheirQValues, double epsilon)

Method Summary

Methods inherited from class burlap.behavior.stochasticgames.agents.maql.MAQLFactory

generateAgent, init

Methods inherited from class java.lang.Object

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

Constructor Detail

MAQLFactory.MAMaxQLearningFactory

public MAQLFactory.MAMaxQLearningFactory(SGDomain d,
                                 double discount,
                                 LearningRate learningRate,
                                 HashableStateFactory hashFactory,
                                 ValueFunctionInitialization qInit,
                                 boolean queryOtherAgentsForTheirQValues,
                                 double epsilon)