Probabilistically Plausible Counterfactual Explanations with Normalizing Flows
[ 1 ] Instytut Informatyki, Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ P ] pracownik
2024
rozdział w monografii naukowej / referat
angielski
- machine learning
- counterfactual explanation
- probabilistic normalized flows
EN We present PPCEF, a novel method for generating prob- abilistically plausible counterfactual explanations (CFs). PPCEF ad- vances beyond existing methods by combining a probabilistic formula- tion that leverages the data distribution with the optimization of plausi- bility within a unified framework. Compared to reference approaches, our method enforces plausibility by directly optimizing the explicit density function without assuming a particular family of parametrized distributions. This ensures CFs are not only valid (i.e., achieve class change) but also align with the underlying data’s probability density. For that purpose, our approach leverages normalizing flows as power- ful density estimators to capture the complex high-dimensional data distribution. Furthermore, we introduce a novel loss function that bal- ances the trade-off between achieving class change and maintaining closeness to the original instance while also incorporating a proba- bilistic plausibility term. PPCEF’s unconstrained formulation allows for an efficient gradient-based optimization with batch processing, leading to orders of magnitude faster computation compared to prior methods. Moreover, the unconstrained formulation of PPCEF allows for the seamless integration of future constraints tailored to specific counterfactual properties. Finally, extensive evaluations demonstrate PPCEF’s superiority in generating high-quality, probabilistically plau- sible counterfactual explanations in high-dimensional tabular settings.
954 - 961
CC BY-NC (uznanie autorstwa - użycie niekomercyjne)
witryna wydawcy
ostateczna wersja opublikowana
w momencie opublikowania
20
140