FastABODDetector<T>
Models & Types · AiDotNet.AnomalyDetection.AngleBased
Detects anomalies using Fast Angle-Based Outlier Detection (FastABOD).
Inheritance: ModelBase<T, Matrix<T>, Vector<T>> → AnomalyDetectorBase<T> → FastABODDetector<T>
Implements: IGeneratedParameterRegistrar<T>
For Beginners
FastABOD is an optimized version of ABOD that uses only k-nearest neighbors instead of all points. This makes it much faster while maintaining good detection quality.
How It Works
The algorithm works by:
- For each point p, find k nearest neighbors
- Compute angles only between neighbor pairs
- Calculate angle variance as outlier score
When to use:
- Large datasets where full ABOD is too slow
- High-dimensional data
- When you need faster runtime with slight accuracy trade-off
Industry Standard Defaults:
- K (neighbors): 10
- Contamination: 0.1 (10%)
Reference: Kriegel, H., et al. (2008). “Angle-Based Outlier Detection in High-dimensional Data.” KDD.
Constructors
| Constructor | Summary |
|---|---|
FastABODDetector(int k, double contamination, int randomSeed) |
Creates a new FastABOD anomaly detector. |
Properties
| Property | Summary |
|---|---|
int K { get; } |
Gets the number of neighbors used for detection. |
Methods
| Method | Summary |
|---|---|
Vector<T> ScoreAnomalies(Matrix<T> X) |
|
void Fit(Matrix<T> X) |