net.sourceforge.jasa.agent.strategy
Class StimuliResponseStrategy

java.lang.Object
  AbstractStrategy
      net.sourceforge.jasa.agent.strategy.AbstractTradingStrategy
          net.sourceforge.jasa.agent.strategy.FixedQuantityStrategyImpl
              net.sourceforge.jasa.agent.strategy.FixedDirectionStrategy
                  net.sourceforge.jasa.agent.strategy.AdaptiveStrategyImpl
                      net.sourceforge.jasa.agent.strategy.DiscreteLearnerStrategy
                          net.sourceforge.jasa.agent.strategy.StimuliResponseStrategy

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable, AdaptiveStrategy, FixedQuantityStrategy, TradingStrategy, MarketEventListener

public class StimuliResponseStrategy
extends DiscreteLearnerStrategy
implements java.io.Serializable

A trading strategy that uses a stimuli-response learning algorithm, such as the Roth-Erev algorithm, to adapt its trading behaviour in successive market rounds by using the agent's profits in the last round as a reward signal.

Field Summary

protected StimuliResponseLearner

learner The learning algorithm to use.

Fields inherited from class net.sourceforge.jasa.agent.strategy.DiscreteLearnerStrategy

markupScale

Fields inherited from class net.sourceforge.jasa.agent.strategy.AbstractTradingStrategy

auction, tradeDirectionPolicy

Constructor Summary

StimuliResponseStrategy()

StimuliResponseStrategy(AbstractTradingAgent agent)

Method Summary

int	act() Generate an action from the learning algorithm.
Learner	getLearner()
void	learn(Market auction) Perform learning.
java.lang.Object	protoClone()
void	reset()
void	setLearner(Learner learner)
java.lang.String	toString()

Methods inherited from class net.sourceforge.jasa.agent.strategy.DiscreteLearnerStrategy

eventOccurred, getMarkupScale, initialise, modifyShout, onRoundFinished, setMarkupScale, subscribeToEvents

Methods inherited from class net.sourceforge.jasa.agent.strategy.FixedDirectionStrategy

isBuy, isSell, setBuy

Methods inherited from class net.sourceforge.jasa.agent.strategy.FixedQuantityStrategyImpl

determineQuantity, getQuantity, setQuantity

Methods inherited from class net.sourceforge.jasa.agent.strategy.AbstractTradingStrategy

clone, getAgent, getTradeDirectionPolicy, isBuy, modifyOrder, setAgent, setTradeDirectionPolicy

Methods inherited from class java.lang.Object

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

Methods inherited from interface net.sourceforge.jasa.agent.TradingStrategy

determineQuantity, isBuy, modifyOrder, setAgent, subscribeToEvents

Field Detail

learner

protected StimuliResponseLearner learner

The learning algorithm to use.

Constructor Detail

StimuliResponseStrategy

public StimuliResponseStrategy(AbstractTradingAgent agent)

StimuliResponseStrategy

public StimuliResponseStrategy()

Method Detail

protoClone

public java.lang.Object protoClone()

Overrides:

protoClone in class AbstractTradingStrategy

act

public int act()

Description copied from class: DiscreteLearnerStrategy

Generate an action from the learning algorithm.

Specified by:

act in class DiscreteLearnerStrategy

learn

public void learn(Market auction)

Description copied from class: DiscreteLearnerStrategy

Perform learning.

Specified by:

learn in class DiscreteLearnerStrategy

reset

public void reset()

Overrides:

reset in class AbstractTradingStrategy

getLearner

public Learner getLearner()

Specified by:

getLearner in interface AdaptiveStrategy

setLearner

public void setLearner(Learner learner)

Specified by:

setLearner in interface AdaptiveStrategy

toString

public java.lang.String toString()