BELLA: Black-box model Explanations by Local Linear Approximations
Abstract: Understanding the decision-making process of black-box models has become not just a legal requirement, but also an additional way to assess their performance. However, the state of the art post-hoc explanation approaches for regression models rely on synthetic data generation, which introduces uncertainty and can hurt the reliability of the explanations. Furthermore, they tend to produce explanations that apply to only very few data points. In this paper, we present BELLA, a deterministic model-agnostic post-hoc approach for explaining the individual predictions of regression black-box models. BELLA provides explanations in the form of a linear model trained in the feature space. BELLA maximizes the size of the neighborhood to which the linear model applies so that the explanations are accurate, simple, general, and robust.
Submission Length: Regular submission (no more than 12 pages of main content)
Previous TMLR Submission Url: https://openreview.net/forum?id=UMi289YXdD
Changes Since Last Submission: Dear editors,
following an encouragement of our previous action editor, we are proposing a resubmission of our work. We have made two major changes:
1) we have addressed all reviewer comments (new text passages are marked in blue), and these changes were acknowledged by the action editor
2) we have made extra-clear in the Limitations section that our method currently does not scale well.
3) we have completely removed the claim that our method can produce counterfactual explanations, which was the main remaining issue that the action editor identified.
We have then shortened the paper to fit into the 12 pages of standard TMLR submissions. We have attached the previous round of reviews in the supplementary material for reference.
**If possible, we would like to request the same action editor, and the same set of reviewers.**
Thank you for your continued support,
the authors
Assigned Action Editor: ~Dennis_Wei1
Submission Number: 4575
Loading