Predictability and Comprehensibility in Post-Hoc XAI Methods: A User-Centered Analysis

Anahid Wachsenegger (Autor:in und Vortragende:r), Haslhofer Bernhard, Simone Kriglstein, Andreas Rauber

Publikation: Beitrag in Buch oder TagungsbandVortrag mit Beitrag in TagungsbandBegutachtung

Abstract

Post-hoc explainability methods aim to clarify predictions of black-box machine learning models. However, it is still largely unclear how well users comprehend the provided explanations and whether these increase the users’ ability to predict the model behavior. We approach this question by conducting a user study to evaluate comprehensibility and predictability in two widely used tools: LIME and SHAP. Moreover, we investigate the effect of counterfactual explanations and misclassifications on users’ ability to understand and predict the model behavior. We find that the comprehensibility of SHAP is significantly reduced when explanations are provided for samples near a model’s decision boundary. Furthermore, we find that counterfactual explanations and misclassifications can significantly increase the users’ understanding of how a machine learning model is making decisions. Based on our findings, we also derive design recommendations for future post-hoc explainability methods with increased comprehensibility and predictability.
OriginalspracheEnglisch
TitelIntelligent Computing. SAI 2023
UntertitelProceedings of the 2023 Computing Conference, Volume 1
Redakteure/-innenKohei Arai
Seiten712-733
ISBN (elektronisch)978-3-031-37717-4
DOIs
PublikationsstatusVeröffentlicht - Sept. 2023
VeranstaltungComputing Conference 2023 - Clayton Hotel Chiswick, London, Großbritannien/Vereinigtes Königreich
Dauer: 22 Juni 202323 Juni 2023

Publikationsreihe

NameLecture Notes in Networks and Systems (LNNS)
Band711

Konferenz

KonferenzComputing Conference 2023
Land/GebietGroßbritannien/Vereinigtes Königreich
StadtLondon
Zeitraum22/06/2323/06/23

Research Field

  • Ehemaliges Research Field - Data Science
  • Former Research Field - Experience Business Transformation

Schlagwörter

  • eXplainable Artificial Intelligence
  • Machine Learning

Fingerprint

Untersuchen Sie die Forschungsthemen von „Predictability and Comprehensibility in Post-Hoc XAI Methods: A User-Centered Analysis“. Zusammen bilden sie einen einzigartigen Fingerprint.

Diese Publikation zitieren