Java* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 1

List of all members
Parameter Class Reference

Base class for parameters of the gradient boosted trees classification training algorithm. More...

Detailed Description

Member Function Documentation

long getFeaturesPerNode ( )

Returns number of features tried as possible splits by the gradient boosted trees classification training algorithm If 0 then sqrt(p) is used, where p is the total number of features.

Returns
Number of features
double getLambda ( )

Returns lambda parameter of the training procedure (L2 regularization on weights lambda) Range: [0, inf). Default is 1

Returns
lambda value
long getMaxIterations ( )

Returns maximal number of iterations parameter of the gradient boosted trees training algorithm

Returns
Number of iterations
long getMaxTreeDepth ( )

Returns maximal tree depth. Default is 0 (unlimited)

Returns
Maximal tree depth
long getMinObservationsInLeafNode ( )

Returns minimal number of samples per node. Default is 1

Returns
Minimal number of samples
double getMinSplitLoss ( )

Returns loss regularization parameter of the training procedure. Min loss reduction required to make a further partition on a leaf node of the tree. Range: [0, inf). Default is 0

Returns
minSplitLoss value
double getObservationsPerTreeFraction ( )

Returns fraction of observations used for a training of one tree, 0 to 1. Default is 1 (sampling without replacement)

Returns
Fraction of observations
double getShrinkage ( )

Returns shrinkage parameter (learning rate) of the training procedure Scales the contribution of each tree by a factor (0, 1]. Default is 0.3

Returns
Shrinkage value
void setEngine ( com.intel.daal.algorithms.engines.BatchBase  engine)

Sets the engine to be used by the algorithm

Parameters
engineto be used by the algorithm
void setFeaturesPerNode ( long  value)

Sets the number of features tried as possible splits by gradient boosted trees classification training algorithm. If 0 then sqrt(p) is used, where p is the total number of features.

Parameters
valueNumber of features
void setLambda ( double  value)

Sets lambda parameter value

Parameters
valuelambda value
void setMaxIterations ( long  n)

Sets maximal number of iterations for the gradient boosted trees training algorithm

Parameters
nNumber of iterations
void setMaxTreeDepth ( long  value)

Sets maximal tree depth. Default is 0 (unlimited)

Parameters
valueMaximal tree depth
void setMinObservationsInLeafNode ( long  value)

Sets minimal number of samples per node. Default is 1

Parameters
valueMinimal number of samples
void setMinSplitLoss ( double  value)

Sets minSplitLoss parameter value

Parameters
valueminSplitLoss value
void setObservationsPerTreeFraction ( double  value)

Sets fraction of observations used for a training of one tree, 0 to 1

Parameters
valueFraction of observations
void setShrinkage ( double  value)

Sets shrinkage parameter value

Parameters
valueShrinkage value

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

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