C++ API Reference for Intel® Data Analytics Acceleration Library 2020 Update 1

List of all members
Parameter Struct Reference

LogitBoost algorithm parameters. More...

Class Declaration

struct DAAL_EXPORT Parameter : public classifier::Parameter
{
Parameter(services::SharedPtr<classifier::training::Batch> wlTrainForParameter,
services::SharedPtr<classifier::prediction::Batch> wlPredictForParameter,
double acc = 0.3, size_t maxIter = 10, double nrAcc = 1.0e-3, size_t nrMaxIter = 100, double dcThreshold = 1.0e-2);
services::SharedPtr<classifier::training::Batch> weakLearnerTraining;
services::SharedPtr<classifier::prediction::Batch> weakLearnerPrediction;
size_t maxIterations;
double newtonRaphsonAccuracyThreshold;
size_t newtonRaphsonMaxIterations;
double degenerateCasesThreshold;
services::Status check() const DAAL_C11_OVERRIDE;
};

Constructor & Destructor Documentation

Parameter ( )

Default constructor

Parameter ( const services::SharedPtr< regression::training::Batch > &  wlTrainForParameter,
const services::SharedPtr< regression::prediction::Batch > &  wlPredictForParameter,
double  acc = 0.0,
size_t  maxIter = 10,
size_t  nC = 0,
double  wThr = 1e-10,
double  zThr = 1e-10 
)

Constructs LogitBoost parameter structure

Parameters
[in]wlTrainForParameterPointer to the training algorithm of the weak learner
[in]wlPredictForParameterPointer to the prediction algorithm of the weak learner
[in]accAccuracy of the LogitBoost training algorithm
[in]maxIterMaximal number of terms in additive regression
[in]nCNumber of classes in the training data set
[in]wThrThreshold to avoid degenerate cases when calculating weights W
[in]zThrThreshold to avoid degenerate cases when calculating responses Z

Member Data Documentation

double accuracyThreshold

Accuracy of the LogitBoost training algorithm

size_t maxIterations

Maximal number of terms in additive regression

double responsesDegenerateCasesThreshold

Threshold to avoid degenerate cases when calculating responses Z

services::SharedPtr<regression::prediction::Batch> weakLearnerPrediction

The algorithm for prediction based on a weak learner model

services::SharedPtr<regression::training::Batch> weakLearnerTraining

The algorithm for weak learner model training

double weightsDegenerateCasesThreshold

Threshold to avoid degenerate cases when calculating weights W


The documentation for this struct was generated from the following file:

For more complete information about compiler optimizations, see our Optimization Notice.