Traverse the trees in a Random Forest ensemble and count the number of times each feature index appears over all the trees
Traverse the trees in a Random Forest ensemble and count the number of times each feature index appears over all the trees
A mapping from feature index, to the number of times it appeared over the trees in the ensemble
Traverse the trees in a Random Forest ensemble and count the number of times each feature appears over all the trees
Traverse the trees in a Random Forest ensemble and count the number of times each feature appears over all the trees
A partial function that returns the name of a feature, given its index
A mapping from feature name to the number of times it appeared over the trees in the ensemble
Traverse the trees in a Random Forest ensemble and count the number of times each feature index appears over all the trees
Traverse the trees in a Random Forest ensemble and count the number of times each feature index appears over all the trees
A sequence of pairs (idx, n) where 'idx' is a feature index and 'n' is the number of times it was used in the ensemble's trees, sorted in descending order of counts.
Traverse the trees in a Random Forest ensemble and count the number of times each feature appears over all the trees
Traverse the trees in a Random Forest ensemble and count the number of times each feature appears over all the trees
A partial function that returns the name of a feature, given its index
A sequence of pairs (name, n) where 'name' is a feature name and 'n' is the number of times it was used in the ensemble's trees, sorted in descending order of counts.
Map an RDD of feature vectors to a corresponding RDD of leaf-node id vectors
Map an RDD of feature vectors to a corresponding RDD of leaf-node id vectors
The RDD of feature vectors
An RDD of leaf-node id vectors
Evaluate an input feature vector and return a vector of the leaf-node ids that the feature vector "landed" at for each tree in the RF ensemble
Evaluate an input feature vector and return a vector of the leaf-node ids that the feature vector "landed" at for each tree in the RF ensemble
The feature vector to evaluate
vector of leaf node ids, one from each tree in the ensemble
Traverse the trees in a Random Forest ensemble and convert each path from root to a leaf into a "rule" that is a sequence of individual predicates representing the decision made at each internal node.
Traverse the trees in a Random Forest ensemble and convert each path from root to a leaf into a "rule" that is a sequence of individual predicates representing the decision made at each internal node.
a partial function that returns name of a feature given its index
a partial function from feature index to number of categories. If an index is not present then it is assumed to be numeric
a map from leaf-node prediction values to a collection of all rules that will yield that value.
Enhance Spark RandomForestModel objects with methods for Random Forest Clustering