Class RealHyp2

Inheritance Relationships

Base Type

Class Documentation

class RealHyp2 : public RealHyp

RealHyp2 is a derived class from RealHyp that overrides the behavior of going left or right in the decision tree.

This class provides specific implementations for the methods to move left and right in the decision tree during the rule extraction process.

Public Functions

RealHyp2(DataSet &data, std::shared_ptr<BpNN> nn, int nbBins, int nbIn, int multiple, int nbWeightLayers)

Constructs a RealHyp2 object using the given dataset and neural network parameters.

Parameters:

  • data – The dataset used for rule extraction.

  • nn – A shared pointer to the neural network.

  • nbBins – The number of bins for discretization.

  • nbIn – The number of input variables.

  • multiple – The multiplicity factor for virtual hyperplanes.

  • nbWeightLayers – The number of weight layers in the neural network.

RealHyp2(std::shared_ptr<VirtualHyp> globalVirt, int nbNets, float *out, int nbOut, DataSet &data, std::shared_ptr<BpNN> nn, int nbBins, int nbIn, int multiple, int nbWeightLayers)

Constructs a RealHyp2 object using the given virtual hyperplanes and neural network parameters.

Parameters:

  • globalVirt – A shared pointer to the global virtual hyperplanes.

  • nbNets – The number of networks.

  • out – The output values from the neural network.

  • nbOut – The number of output variables.

  • data – The dataset used for rule extraction.

  • nn – A shared pointer to the neural network.

  • nbBins – The number of bins for discretization.

  • nbIn – The number of input variables.

  • multiple – The multiplicity factor for virtual hyperplanes.

  • nbWeightLayers – The number of weight layers in the neural network.