Source

Extra Tree Regressor#

An Extremely Randomized Regression Tree. These trees differ from standard Regression Trees in that they choose a split drawn completely at random, rather than searching the entire column for the best split. Due to random splitting, Extra Trees are faster to build and their predictions have higher variance than a regular decision tree.

Interfaces: Estimator, Learner, Persistable

Data Type Compatibility: Categorical, Continuous

Parameters#

# Param Default Type Description
1 max depth PHP_INT_MAX int The maximum depth of a branch in the tree.
2 max leaf size 3 int The max number of samples that a leaf node can contain.
3 max features Auto int The number of features to consider when determining a best split.
4 min purity increase 1e-7 float The minimum increase in purity necessary for a node not to be post pruned.

Additional Methods#

Return the normalized feature importances i.e. the proportion that each feature contributes to the overall model, indexed by feature column:

public featureImportances() : array

Display a human readable text representation of the decision tree:

public printrules() : void

Return the height of the tree:

public height() : int

Return the balance factor of the tree:

public balance() : int

Example#

use Rubix\ML\Classifiers\ExtraTreeRegressor;

$estimator = new ExtraTreeRegressor(30, 3, 20, 0.05);

References#

  • P. Geurts et al. (2005). Extremely Randomized Trees.