We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Diverse, Global and Amortised Counterfactual Explanations for Uncertainty Estimates

Abstract: To interpret uncertainty estimates from differentiable probabilistic models, recent work has proposed generating a single Counterfactual Latent Uncertainty Explanation (CLUE) for a given data point where the model is uncertain, identifying a single, on-manifold change to the input such that the model becomes more certain in its prediction. We broaden the exploration to examine $\delta$-CLUE, the set of potential CLUEs within a $\delta$ ball of the original input in latent space. We study the diversity of such sets and find that many CLUEs are redundant; as such, we propose DIVerse CLUE ($\nabla$-CLUE), a set of CLUEs which each propose a distinct explanation as to how one can decrease the uncertainty associated with an input. We then further propose GLobal AMortised CLUE (GLAM-CLUE), a distinct and novel method which learns amortised mappings on specific groups of uncertain inputs, taking them and efficiently transforming them in a single function call into inputs for which a model will be certain. Our experiments show that $\delta$-CLUE, $\nabla$-CLUE, and GLAM-CLUE all address shortcomings of CLUE and provide beneficial explanations of uncertainty estimates to practitioners.
Comments: Accepted as a conference paper to AAAI 2022
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (stat.ML)
Cite as: arXiv:2112.02646 [cs.LG]
  (or arXiv:2112.02646v3 [cs.LG] for this version)

Submission history

From: Dan Ley [view email]
[v1] Sun, 5 Dec 2021 18:27:21 GMT (24562kb,D)
[v2] Tue, 7 Dec 2021 10:38:43 GMT (24561kb,D)
[v3] Thu, 9 Dec 2021 01:48:15 GMT (24561kb,D)

Link back to: arXiv, form interface, contact.