LIME (Local Interpretable Model-agnostic Explanations)

Explanation Function

Mathematical function that, for a given instance, returns an interpretable explanation by minimizing the fidelity loss between the original model and the surrogate.

← Tillbaka