IModelExplainer<T>
Interfaces · AiDotNet.Interfaces
Interface for model-agnostic explainers that can explain any predictive model’s decisions.
Known implementations: IGlobalExplainer<T, TExplanation>, ILocalExplainer<T, TExplanation>
For Beginners
Model explainers help you understand WHY a model makes certain predictions. Unlike models that implement IInterpretableModel directly, these explainers work with ANY model - they treat the model as a “black box” and analyze its behavior by observing inputs and outputs.
Think of it like understanding how a vending machine works: you don’t need to see inside it, you just try different button combinations and observe what comes out.
Properties
| Property | Summary |
|---|---|
bool SupportsGlobalExplanations { get; } |
Gets whether this explainer provides global (model-wide) explanations. |
bool SupportsLocalExplanations { get; } |
Gets whether this explainer provides local (per-instance) explanations. |
string MethodName { get; } |
Gets the name of this explanation method. |