Random Forest Ensemble

GPTKB entity

Statements (49)
Predicate Object
gptkbp:instanceOf gptkb:model
ensemble learning method
gptkbp:advantage parallelizable
can be slow with many trees
large memory usage
handles large datasets
less interpretable than single trees
gptkbp:canBe binary classification
feature selection
anomaly detection
multi-class classification
regression tasks
ensemble stacking
imputation of missing values
ranking feature importance
gptkbp:category supervised learning
gptkbp:citation gptkb:Breiman,_L._(2001)._Random_Forests._Machine_Learning,_45(1),_5-32.
gptkbp:developedBy gptkb:Leo_Breiman
gptkbp:featureImportance can be estimated
gptkbp:handles high-dimensional data
missing values
https://www.w3.org/2000/01/rdf-schema#label Random Forest Ensemble
gptkbp:hyperparameter maximum tree depth
minimum samples per leaf
number of features per split
number of trees
gptkbp:implementedIn gptkb:Spark_MLlib
gptkb:Weka
gptkb:scikit-learn
R
gptkbp:improves prediction accuracy
gptkbp:introducedIn 2001
gptkbp:output average prediction (regression)
majority vote (classification)
gptkbp:reduces overfitting
gptkbp:relatedTo gptkb:Gradient_Boosting
gptkb:AdaBoost
bagging
boosting
Extra Trees
gptkbp:robustTo outliers
gptkbp:technique bagging
gptkbp:usedFor gptkb:dictionary
regression
gptkbp:uses decision trees
random feature selection
bootstrap sampling
gptkbp:bfsParent gptkb:RFE
gptkbp:bfsLayer 7