Actionable counterfactual explanations using Bayesian networks and path planning with applications to environmental quality improvement

Enrique Valero-Leal, Pedro Larrañaga, Concha Bielza · 2025

Read Actionable counterfactual explanations using Bayesian networks and path planning with applications to environmental quality improvement by Enrique Valero-Leal, Pedro Larrañaga, Concha Bielza (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-02634.txt.

Description

A good explanation of a model's verdict tells the user what to change to get a different one, but only changes a person can actually make count. Enrique Valero-Leal and two co-authors build such explanations in 2025 without peeking at the raw training data, which may be private. They learn a Bayesian network as a landscape, then plan a path across it from the real case to a better one.

On fifteen test sets their paths were simpler and more doable than those from leading rivals. A run on environmental agency data linked county quality of life to housing strain in ways that protect fairness across neighborhoods. The authors say the network's clarity suits high stakes decisions. Every suggested change, they stress, must leave the person better off overall.

Catalog record

CREATOREnrique Valero-Leal, Pedro Larrañaga, Concha Bielza
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE73 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-02634

Browse the full library · About the Tower of Babel