Prototype-based Explanations

Nearest Neighbor Explanation

Interpretability technique that explains a model's decision by identifying and presenting the training examples closest to the instance to be explained.

← Tillbaka