Viewing the process of generating counterfactuals as a source of knowledge: a new approach for explaining classifiers

2023/09/08 by Vincent Lemaire, Lemaire, Vincent, Nathan Le Boudec +5 · 1 citation
Computer Science · #Bayesian Modeling and Causal Inference #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Imbalanced Data Classification Techniques #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2309.04284

openalex publication_date 2023/09/08 · openalex created_date 2023/09/12 · openalex updated_date 2026/07/28

Abstract

There are now many explainable AI methods for understanding the decisions of a machine learning model. Among these are those based on counterfactual reasoning, which involve simulating features changes and observing the impact on the prediction. This article proposes to view this simulation process as a source of creating a certain amount of knowledge that can be stored to be used, later, in different ways. This process is illustrated in the additive model and, more specifically, in the case of the naive Bayes classifier, whose interesting properties for this purpose are shown.

Cited by

• Khiops: An End-to-End, Frugal AutoML and XAI Machine Learning Solution for Large, Multi-Table Databases

Related

• Counterfactual Instances Explain Little
• On the computation of counterfactual explanations -- A survey
• Explaining Data-Driven Decisions made by AI Systems: The Counterfactual Approach
• Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers
• Relevant Irrelevance: Generating Alterfactual Explanations for Image Classifiers