org.vizzini.game.boardgame
Class AbstractNeuralGridBoardEvaluator

java.lang.Object
  org.vizzini.game.AbstractEvaluator
      org.vizzini.game.AbstractNeuralNetworkEvaluator
          org.vizzini.game.boardgame.AbstractNeuralGridBoardEvaluator

All Implemented Interfaces:

Serializable, IEvaluator

Direct Known Subclasses:

NeuralNetworkEvaluator

public abstract class AbstractNeuralGridBoardEvaluator
extends AbstractNeuralNetworkEvaluator

Provides an evaluator which uses a neural network to arrive at a grid board evaluation.

Since:

v0.1

Version:

v0.2

Author:

Jeffrey M. Thompson

See Also:

Serialized Form

Constructor Summary

AbstractNeuralGridBoardEvaluator()

Method Summary

protected String	arrayToString(double[] array) Return a string representation of the given array.
protected abstract double[]	encode(IGridBoard gridBoard, IAgent agent) Encode the given grid board as appropriate for a neural network input.
protected int	evaluateTokens(IEnvironment environment, IAgent agent) Evaluate the tokens in the given environment from the perspective of the given agent.

Methods inherited from class org.vizzini.game.AbstractNeuralNetworkEvaluator

configure, getNeuralNetwork, setNeuralNetwork

Methods inherited from class org.vizzini.game.AbstractEvaluator

evaluate, multiplier

Methods inherited from class java.lang.Object

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

Constructor Detail

AbstractNeuralGridBoardEvaluator

public AbstractNeuralGridBoardEvaluator()

Method Detail

encode

protected abstract double[] encode(IGridBoard gridBoard,
                                   IAgent agent)

Encode the given grid board as appropriate for a neural network input.

Since:

v0.1

arrayToString

protected String arrayToString(double[] array)

Return a string representation of the given array.

Since:

v0.1

evaluateTokens

protected int evaluateTokens(IEnvironment environment,
                             IAgent agent)

Evaluate the tokens in the given environment from the perspective of the given agent.

Specified by:

evaluateTokens in class AbstractEvaluator

Parameters:

environment - The environment to evaluate.

agent - The agent of interest.

Since:

v0.1