gcluster.h

一个由Mike Gashler完成的机器学习方面的includes neural net, naive bayesian classifier, decision tree, KNN, a genet

/*
	Copyright (C) 2006, Mike Gashler

	This library is free software; you can redistribute it and/or
	modify it under the terms of the GNU Lesser General Public
	License as published by the Free Software Foundation; either
	version 2.1 of the License, or (at your option) any later version.

	see http://www.gnu.org/copyleft/lesser.html
*/

#ifndef __GCLUSTER_H__
#define __GCLUSTER_H__

class GIntArray;
class GArffRelation;
class GArffData;

// This one is slow. It finds the best matches and merges them until the desired
// number of clusters is found.
class GMergeClusterer
{
protected:
	GArffRelation* m_pRelation;
	GArffData* m_pData;
	int m_nClusters;
	int* m_pClusters;

public:
	GMergeClusterer(GArffRelation* pRelation, GArffData* pData, int nClusters);
	~GMergeClusterer();

	int GetCluster(int nVector);

protected:
	bool FindBestMatch(int* pA, int* pB);
	void ChangeCluster(int nFrom, int nTo);
};


// An implementation of the K-means clustering algorithm. Note that this operates
// only on input attributes. If you want it to consider all attributes, set them
// all to be input attributes before constructing this class.
class GKMeans
{
protected:
	GArffRelation* m_pRelation;
	GArffData* m_pData;
	int m_nClusters;
	int* m_pClusters;

public:
	GKMeans(GArffRelation* pRelation, GArffData* pData, int nClusters);
	~GKMeans();

	int GetCluster(int nVector);

protected:
	bool Cluster(int nMaxIterations);
	bool SelectSeeds(GArffData* pSeeds);
};



// This clusters every vector with it's k nearest neighbors. It's somewhat
// fast, but it isn't very noise tolerant.
class GNeighborClusterer
{
protected:
	GIntArray* m_pClusters;
	GIntArray* m_pVectors;

	GNeighborClusterer(int nVectorCount);
public:
	static GNeighborClusterer* Cluster(GArffRelation* pRelation, GArffData* pData, int nNeighbors);
	~GNeighborClusterer();

	int GetClusterCount();
	int GetClusterSize(int nCluster);
	int GetVectorIndex(int nCluster, int nVector);

#ifndef NO_TEST_CODE
	static void Test();
#endif // NO_TEST_CODE
};

#endif // __GCLUSTER_H__