net.sourceforge.cilib.games.game.predatorprey
Class NeuralPredatorOutputStrategy

java.lang.Object
  net.sourceforge.cilib.games.agent.neural.NeuralOutputInterpretationStrategy
      net.sourceforge.cilib.games.game.predatorprey.NeuralPredatorOutputStrategy

public class NeuralPredatorOutputStrategy
extends NeuralOutputInterpretationStrategy

Author:

leo This is a NeuralOutputInterpretationStrategy that will interperet the output of a neural network to make a movement decision for a Predator agent.

Constructor Summary

NeuralPredatorOutputStrategy()

Method Summary

void	applyOutputToState(TypeList outputData, Agent currentPlayer, Game oldState) Modify the game by the output vector
int	getAmOutputs() This strategy requires 4 outputs: The first determines movement on the x axis the second determines if that movement is left or right The third determines movement on the y axes The fourth determines if it is up or down

Methods inherited from class java.lang.Object

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

Constructor Detail

NeuralPredatorOutputStrategy

public NeuralPredatorOutputStrategy()

Method Detail

applyOutputToState

public void applyOutputToState(TypeList outputData,
                               Agent currentPlayer,
                               Game oldState)

Modify the game by the output vector

Specified by:

applyOutputToState in class NeuralOutputInterpretationStrategy

Parameters:

outputData - the output data vector

currentPlayer - the agent that is represented by the Neural Network

oldState - the game state to be altered

getAmOutputs

public int getAmOutputs()

This strategy requires 4 outputs: The first determines movement on the x axis the second determines if that movement is left or right The third determines movement on the y axes The fourth determines if it is up or down

Specified by:

getAmOutputs in class NeuralOutputInterpretationStrategy

Returns:

the output count