Base class for parameters of the gradient boosted trees regression training algorithm.
More...
long getFeaturesPerNode |
( |
| ) |
|
Returns number of features tried as possible splits by the gradient boosted trees regression training algorithm If 0 then p/3 is used, where p is the total number of features.
- Returns
- Number of features
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
Returns maximal tree depth. Default is 0 (unlimited)
- Returns
- Maximal tree depth
long getMinObservationsInLeafNode |
( |
| ) |
|
Returns minimal number of samples per node. Default is 5
- 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 with replacement)
- Returns
- Fraction of observations
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
Sets the engine to be used by the algorithm
- Parameters
-
engine | to be used by the algorithm |
void setFeaturesPerNode |
( |
long |
value | ) |
|
Sets the number of features tried as possible splits by gradient boosted trees regression training algorithm. If 0 then p/3 is used, where p is the total number of features.
- Parameters
-
void setLambda |
( |
double |
value | ) |
|
Sets lambda parameter value
- Parameters
-
void setMaxIterations |
( |
long |
n | ) |
|
Returns maximal number of iterations parameter of the gradient boosted trees training algorithm
- Returns
- Number of iterations
void setMaxTreeDepth |
( |
long |
value | ) |
|
Sets maximal tree depth. Default is 0 (unlimited)
- Parameters
-
void setMinObservationsInLeafNode |
( |
long |
value | ) |
|
Sets minimal number of samples per node. Default is 5
- Parameters
-
value | Minimal number of samples |
void setMinSplitLoss |
( |
double |
value | ) |
|
Sets minSplitLoss parameter value
- Parameters
-
void setObservationsPerTreeFraction |
( |
double |
value | ) |
|
Sets fraction of observations used for a training of one tree, 0 to 1
- Parameters
-
value | Fraction of observations |
void setShrinkage |
( |
double |
value | ) |
|
Sets shrinkage parameter value
- Parameters
-
The documentation for this class was generated from the following file:
- gbt/regression/training/Parameter.java