Statements (50)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
Machine Learning Algorithm
|
| gptkbp:advantage |
Computationally Intensive
Sensitive to Hyperparameters Handles Mixed Data Types High Predictive Accuracy Less Interpretable |
| gptkbp:alternativeTo |
gptkb:Neural_Networks
gptkb:Random_Forests gptkb:Support_Vector_Machines |
| gptkbp:application |
Medical Diagnosis
Fraud Detection Credit Scoring Click-Through Rate Prediction Customer Churn Prediction Ranking Problems |
| gptkbp:basedOn |
Decision Trees
|
| gptkbp:canBe |
Categorical Variables
Missing Values |
| gptkbp:category |
gptkb:Gradient_Boosting
Supervised Learning Tree-Based Methods |
| gptkbp:feature |
Handles Non-linear Data
Prone to Overfitting Reduces Bias Sequentially Adds Trees Supports Feature Importance |
| https://www.w3.org/2000/01/rdf-schema#label |
Gradient Boosted Trees
|
| gptkbp:hyperparameter |
Learning Rate
Number of Trees Subsample Ratio Tree Depth Loss Function |
| gptkbp:implementedIn |
gptkb:CatBoost
gptkb:LightGBM gptkb:XGBoost gptkb:scikit-learn |
| gptkbp:introduced |
gptkb:Jerome_H._Friedman
|
| gptkbp:introducedIn |
1999
|
| gptkbp:lossFunction |
gptkb:Mean_Squared_Error
Huber Loss Log Loss |
| gptkbp:relatedTo |
gptkb:AdaBoost
gptkb:Random_Forests Boosting Ensemble Learning |
| gptkbp:requires |
Labeled Data
|
| gptkbp:usedFor |
Regression
Classification |
| gptkbp:bfsParent |
gptkb:C&RT
|
| gptkbp:bfsLayer |
7
|