Anchors: High Precision Model-Agnostic Explanations

← Back to document page

Anchors: High- Precision Model-Agnostic ExplanationsMarco Tulio RibeiroUniversity of SinghUniversity of California, GuestrinUniversity of introduce a novel Model-Agnostic system that explains thebehavior of complex models with high- Precision rules calledanchors, representing local, sufficient conditions for predic-tions. We propose an algorithm to efficiently compute theseexplanations for any black-box model with high-probabilityguarantees. We demonstrate the flexibility of anchors by ex-plaining a myriad of different models for different domainsand tasks.

coverage to the model’s behavior (the anchor on the right of Figure 2b is broader) and making their boundaries clear. Leaving the discussion of how to compute anchors for later, we now demonstrate their usefulness and flexibility via concrete examples in a variety of domains and models. Text Classification: We have already alluded to Figure 1,

  Concrete, Anchor