EnsembleClassifierBase<T>
Base Classes · AiDotNet.Classification.Ensemble
Base class for ensemble classification methods that combine multiple classifiers.
Inheritance: ClassifierBase<T> → ProbabilisticClassifierBase<T> → EnsembleClassifierBase<T>
Implements: IGeneratedParameterRegistrar<T>
Known derived types: AdaBoostClassifier<T>, DARTClassifier<T>, ExplainableBoostingClassifier<T>, ExtraTreesClassifier<T>, GradientBoostingClassifier<T>, NGBoostClassifier<T>, RandomForestClassifier<T>
For Beginners
Ensemble learning is like getting opinions from a group of experts instead of just one.
Imagine you want to predict if a movie will be successful. You could:
- Ask just one expert (single classifier)
- Ask many experts and combine their opinions (ensemble)
The second approach is usually more reliable because:
- Individual experts may have blind spots that others don’t
- Combining diverse opinions often leads to better decisions
- Errors from one expert may be corrected by others
Common ensemble strategies:
- Bagging: Train on different random subsets of data
- Boosting: Train sequentially, focusing on mistakes
- Voting: Let classifiers vote on the answer
How It Works
Ensemble methods combine multiple individual classifiers (base estimators) to produce a more robust and accurate prediction than any single classifier could achieve.
Properties
| Property | Summary |
|---|---|
Vector<T> FeatureImportances { get; set; } |
Gets or sets the feature importances aggregated across all estimators. |
int NEstimators { get; } |
The number of estimators in the ensemble. |
Methods
| Method | Summary |
|---|---|
Matrix<T> PredictProbabilities(Matrix<T> input) |
Aggregates predictions from all estimators in the ensemble. |
ModelMetadata<T> GetModelMetadata() |