IsolationForest<T>
Models & Types · AiDotNet.AnomalyDetection.TreeBased
Implements the Isolation Forest algorithm for anomaly detection.
Inheritance: ModelBase<T, Matrix<T>, Vector<T>> → AnomalyDetectorBase<T> → IsolationForest<T>
For Beginners
Isolation Forest is an efficient algorithm for detecting anomalies. The key insight is that anomalies are “few and different” - they are easier to isolate from the rest of the data.
How It Works
The algorithm works by:
- Building a “forest” of random isolation trees
- Each tree randomly partitions the data by selecting random features and split points
- Anomalies tend to be isolated in fewer steps (closer to the root of the tree)
- Normal points require more splits to isolate (deeper in the tree)
When to use: Isolation Forest is particularly effective for:
- High-dimensional data
- Large datasets (it scales linearly)
- When you don’t know the distribution of your data
Industry Standard Defaults:
- Number of trees: 100 (provides stable results)
- Max samples: 256 (balances speed and accuracy)
- Contamination: 0.1 (10%)
Reference: Liu, F. T., Ting, K. M., and Zhou, Z. H. (2008). “Isolation Forest.” In: ICDM ’08. Eighth IEEE International Conference on Data Mining.
Constructors
| Constructor | Summary |
|---|---|
IsolationForest(int numTrees, int maxSamples, double contamination, int randomSeed) |
Creates a new Isolation Forest anomaly detector. |
Properties
| Property | Summary |
|---|---|
int MaxSamples { get; } |
Gets the number of samples used to train each tree. |
int NumTrees { get; } |
Gets the number of trees in the forest. |
Methods
| Method | Summary |
|---|---|
Vector<T> ScoreAnomalies(Matrix<T> X) |
|
void Fit(Matrix<T> X) |