On Quantitative Evaluations of Counterfactuals

Abstract

As counterfactual examples become increasingly popular for explaining decisions of deep learning models, it is essential to understand what properties quantitative evaluation metrics do capture and equally important what they do not capture. Currently, such understanding is lacking, potentially slowing down scientific progress. In this paper, we consolidate the work on evaluating visual counterfactual examples through an analysis and experiments. We find that while most metrics behave as intended for sufficiently simple datasets, some fail to tell the difference between good and bad counterfactuals when the complexity increases. We observe experimentally that metrics give good scores to tiny adversarial-like changes, wrongly identifying such changes as superior counterfactual examples. To mitigate this issue, we propose two new metrics, the Label Variation Score and the Oracle score, which are both less vulnerable to such tiny changes. We conclude that a proper quantitative evaluation of visual counterfactual examples should combine metrics to ensure that all aspects of good counterfactuals are quantified.

References in corpus (9)

• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• Explaining NonLinear Classification Decisions with Deep Taylor Decomposition
• Explaining Machine Learning Classifiers through Diverse Counterfactual Explanations
• Towards Robust Interpretability with Self-Explaining Neural Networks
• Learning Model-Agnostic Counterfactual Explanations for Tabular Data
• Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers
• Explaining Image Classifiers by Counterfactual Generation
• Counterfactual Explanation Based on Gradual Construction for Deep Networks
• ECINN: Efficient Counterfactuals from Invertible Neural Networks