MetaClassifierBase<T>
Base Classes · AiDotNet.Classification.Meta
Base class for meta classifiers that wrap other classifiers.
Inheritance: ClassifierBase<T> → ProbabilisticClassifierBase<T> → MetaClassifierBase<T>
Known derived types: BaggingClassifier<T>, ClassifierChain<T>, MultiOutputClassifier<T>, OneVsOneClassifier<T>, OneVsRestClassifier<T>, StackingClassifier<T>, VotingClassifier<T>
For Beginners
Meta classifiers are “classifiers about classifiers”:
Examples:
- OneVsRest: Trains one classifier per class
- OneVsOne: Trains one classifier per pair of classes
- Voting: Combines predictions from multiple classifiers
- Stacking: Uses classifier outputs as features for another classifier
They extend the capabilities of base classifiers.
How It Works
Meta classifiers use other classifiers as base estimators to provide enhanced functionality like multi-class support, multi-label support, or ensemble voting.
Properties
| Property | Summary |
|---|---|
bool SupportsParameterInitialization { get; } |
Methods
| Method | Summary |
|---|---|
IFullModel<T, Matrix<T>, Vector<T>> WithParameters(Vector<T> parameters) |