org.apache.spark.streamdm.classifiers
Basic majority class active learning node for Hoeffding tree
ConditionalTest is an abstract class for conditional tests, used for splitting nodes in Hoeffding trees.
Trait FeatureClassObserver for observing the class distribution of one feature.
Class for recording a split suggestion.
The container of a node.
Gaussian incremental estimator that uses incremental method, more resilient to floating point imprecision.
Class GuassianNumericFeatureClassObserver for observing the class data distribution for a numeric feature using gaussian estimators.
Class for computing splitting criteria using Gini with respect to distributions of class values.
The Hoeffding tree is an incremental decision tree learner for large data streams, that assumes that the data distribution is not changing over time.
Inactive learning node for Hoeffding trees
Class for computing splitting criteria using information gain with respect to distributions of class values.
Learning node class type for Hoeffding trees.
Naive Bayes based learning node.
Adaptive Naive Bayes learning node.
Abstract class containing the node information for the Hoeffding trees.
Nominal binary conditional test for splitting nodes in Hoeffding trees.
Class for observing the class distribution of a nominal feature.
Nominal multi-way conditional test for splitting nodes in Hoeffding trees.
NullFeatureClassObserver is a null class for observers.
Numeric binary rule predicate test for splitting nodes in Hoeffding trees.
Numeric binary conditional test for splitting nodes in Hoeffding trees.
Trait for the numeric feature observers.
Trait for computing splitting criteria with respect to distributions of class values.
Branch node of the Hoeffding tree.
Class for computing splitting criteria using variance reduction with respect to distributions of class values.
ojbect FeatureClassObserver for create FeatureClassObserver.