Split Criterions#

Decision trees are built by splitting the data into groups based on a split criterion. The split criterion is a function that measures the quality of a split.

Module containing split criteria for decision trees.

class capymoa.splitcriteria.SplitCriterion[source]#

Split criteria are used to evaluate the quality of a split in a decision tree.

java_object() SplitCriterion[source]#

Return the Java object that this class wraps.

class capymoa.splitcriteria.VarianceReductionSplitCriterion[source]#

Goodness of split criterion based on variance reduction.

__init__()[source]#
java_object() SplitCriterion[source]#

Return the Java object that this class wraps.

class capymoa.splitcriteria.InfoGainSplitCriterion[source]#

Goodness of split using information gain.

__init__(min_branch_frac: float = 0.01)[source]#

Construct InfoGainSplitCriterion.

Parameters:

min_branch_frac – Minimum fraction of weight required down at least two branches.

java_object() SplitCriterion[source]#

Return the Java object that this class wraps.

class capymoa.splitcriteria.GiniSplitCriterion[source]#

Goodness of split using Gini impurity.

__init__()[source]#
java_object() SplitCriterion[source]#

Return the Java object that this class wraps.