RandomForests.h

#ifndef GRT_RANDOM_FORESTS_HEADER
#define GRT_RANDOM_FORESTS_HEADER

#include "../DecisionTree/DecisionTree.h"

GRT_BEGIN_NAMESPACE

class RandomForests : public Classifier
{
public:
 RandomForests(const DecisionTreeNode &decisionTreeNode = DecisionTreeClusterNode(),
                  const UINT forestSize=10,
                  const UINT numRandomSplits=100,
                  const UINT minNumSamplesPerNode=5,
                  const UINT maxDepth=10,
                  const UINT trainingMode = DecisionTree::BEST_RANDOM_SPLIT,
                  const bool removeFeaturesAtEachSpilt = true,
                  const bool useScaling=false,
                  const Float bootstrappedDatasetWeight = 0.8);
    
    RandomForests(const RandomForests &rhs);
    
 virtual ~RandomForests(void);
    
 RandomForests &operator=(const RandomForests &rhs);
    
 virtual bool deepCopyFrom(const Classifier *classifier);
    
    virtual bool train_(ClassificationData &trainingData);
    
    virtual bool predict_(VectorDouble &inputVector);
    
    virtual bool clear();
    
    virtual bool print() const;
    
    virtual bool saveModelToFile( std::fstream &file ) const;
    
    virtual bool loadModelFromFile( std::fstream &file );

    bool combineModels( const RandomForests &forest );
    
    UINT getForestSize() const;
    
    UINT getNumRandomSplits() const;
    
    UINT getMinNumSamplesPerNode() const;
    
    UINT getMaxDepth() const;
    
    UINT getTrainingMode() const;
    
    const Vector< DecisionTreeNode* > &getForest() const;
    
    bool getRemoveFeaturesAtEachSpilt() const;

    Float getBootstrappedDatasetWeight() const;

    DecisionTreeNode* getTree( const UINT index ) const;

    DecisionTreeNode* deepCopyDecisionTreeNode() const;

    VectorDouble getFeatureWeights( const bool normWeights = true ) const;

    MatrixDouble getLeafNodeFeatureWeights( const bool normWeights = true ) const;
    
    bool setForestSize(const UINT forestSize);
    
    bool setNumRandomSplits(const UINT numSplittingSteps);
    
    bool setMinNumSamplesPerNode(const UINT minNumSamplesPerNode);
    
    bool setMaxDepth(const UINT maxDepth);
    
    bool setRemoveFeaturesAtEachSpilt(const bool removeFeaturesAtEachSpilt);
    
    bool setTrainingMode(const UINT trainingMode);
    
    bool setDecisionTreeNode( const DecisionTreeNode &node );

    bool setBootstrappedDatasetWeight( const Float bootstrappedDatasetWeight );
    
    //Tell the compiler we are using the base class train method to stop hidden virtual function warnings
    using MLBase::saveModelToFile;
    using MLBase::loadModelFromFile;
    
protected:
    
    UINT forestSize;
    UINT numRandomSplits;
    UINT minNumSamplesPerNode;
    UINT maxDepth;
    UINT trainingMode;
    bool removeFeaturesAtEachSpilt;
    Float bootstrappedDatasetWeight;
    DecisionTreeNode* decisionTreeNode;
    Vector< DecisionTreeNode* > forest;
    
private:
    static RegisterClassifierModule< RandomForests > registerModule;
    
};

GRT_END_NAMESPACE

#endif //GRT_RANDOM_FORESTS_HEADER