CNeuralNetwork Class Reference

This class provides two neural networks based on a game board. More...

#include <NeuralNetwork.h>

Collaboration diagram for CNeuralNetwork:

Collaboration graph
[legend]
List of all members.

Public Member Functions

double propagate (char *pBoard, char player, bool createTheInputVector=true)
 Gives the output value of the neural network with the given game board as input pattern.
void trainValue (char *pBoard, char player, double trainValue, double learnRate, double maxError)
 Trains the given value to a given game board as input pattern.
int getBestMove (char *pBoard, char player)
 Returns the best rated move based on the given board.
void learn ()
 Trains the networks through competition, with before initialized parameters.
void learnTest ()
 Trains the networks through competition, with before initialized parameters.
void learnTest2 ()
 Trains the networks through competition, with before initialized parameters.
void setLearnParameters ()
 Sets the learn parameters for the networks for a new training session.
bool saveNet (const wchar_t *pFileName)
 Stores the current networks in a file.
bool loadNet (const wchar_t *pFileName)
 Reads the current networks from a file.

Detailed Description

This class provides two neural networks based on a game board.


The documentation for this class was generated from the following files:
Generated on Mon Jan 22 02:40:31 2007 for muele by  doxygen 1.4.7