ExplainableBoostingClassifier<T>
Models & Types · AiDotNet.Classification.Boosting
Explainable Boosting Machine (EBM) for interpretable classification.
Inheritance: ClassifierBase<T> → ProbabilisticClassifierBase<T> → EnsembleClassifierBase<T> → ExplainableBoostingClassifier<T>
For Beginners
EBM is special because it gives you the best of both worlds:
- High accuracy (comparable to gradient boosting and random forests)
- Full interpretability (you can see exactly why each prediction was made)
How it works:
- For each feature, EBM learns a “shape function” that shows how that feature
affects the prediction
- The final prediction is simply the sum of all these shape functions plus
an intercept, passed through sigmoid for probability
- You can plot these shape functions to understand exactly how the model
uses each feature
For example, in predicting loan defaults:
- The shape function for “income” might show higher income = lower risk
- The shape function for “debt_ratio” might show higher ratio = higher risk
- The prediction combines: intercept + f(income) + f(debt_ratio) + …
This additive structure makes EBM uniquely interpretable while still being accurate.
How It Works
EBM is a Generalized Additive Model (GAM) with boosting that provides glass-box interpretability while maintaining high accuracy. It learns smooth functions for each feature and optionally pairwise interactions.
Reference: Lou, Y., et al. “Intelligible Models for Healthcare: Predicting Pneumonia Risk and Hospital 30-day Readmission” (2012).
Example
// Create EBM classifier for interpretable predictions
var options = new ExplainableBoostingClassifierOptions<double>();
// Prepare training data: 6 samples with 2 features
var features = new Matrix<double>(6, 2);
features[0, 0] = 1.0; features[0, 1] = 1.1;
features[1, 0] = 1.2; features[1, 1] = 0.9;
features[2, 0] = 0.8; features[2, 1] = 1.0;
features[3, 0] = 5.0; features[3, 1] = 5.1;
features[4, 0] = 5.2; features[4, 1] = 4.9;
features[5, 0] = 4.8; features[5, 1] = 5.0;
var labels = new Vector<double>(new double[] { 0, 0, 0, 1, 1, 1 });
// Train additive model with per-feature shape functions
var result = new AiModelBuilder<double, Matrix<double>, Vector<double>>()
.ConfigureModel(new ExplainableBoostingClassifier<double>(options))
.Build(features, labels);
// Predict with full interpretability (shape function contributions visible)
var newSample = new Matrix<double>(1, 2);
newSample[0, 0] = 1.1; newSample[0, 1] = 1.0;
var prediction = result.Predict(newSample);
// Result is available in the returned value
Constructors
| Constructor | Summary |
|---|---|
ExplainableBoostingClassifier(ExplainableBoostingClassifierOptions<T> options, IRegularization<T, Matrix<T>, Vector<T>> regularization) |
Initializes a new instance of ExplainableBoostingClassifier. |
Properties
| Property | Summary |
|---|---|
IReadOnlyDictionary<ValueTuple<int, int>, Matrix<T>> InteractionTerms { get; } |
Gets the interaction terms. |
IReadOnlyList<Vector<T>> BinEdges { get; } |
Gets the bin edges for each feature. |
IReadOnlyList<Vector<T>> ShapeFunctions { get; } |
Gets the shape functions for each feature. |
T Intercept { get; } |
Gets the intercept (baseline log-odds). |
Methods
| Method | Summary |
|---|---|
Dictionary<string, T> ExplainPrediction(Vector<T> sample) |
Explains a single prediction by showing each feature’s contribution. |
Matrix<T> PredictProbabilities(Matrix<T> input) |
Predicts class probabilities for input samples. |
ModelMetadata<T> GetModelMetadata() |
|
Vector<T> Predict(Matrix<T> input) |
Predicts class labels for input samples. |
void Train(Matrix<T> x, Vector<T> y) |
Returns the model type identifier. |