Individualised Counterfactual Examples Using Conformal Prediction Intervals

TL;DR

This paper introduces a method for generating individualized counterfactual explanations for black-box classifiers by leveraging conformal prediction intervals to identify the most informative counterfactuals based on the individual's knowledge and prediction uncertainty.

Contribution

It proposes a novel approach combining conformal prediction intervals with counterfactual explanations to personalize and improve the informativeness of counterfactuals for decision understanding.

Findings

Counterfactuals in regions of high uncertainty are more informative.

Synthetic and real datasets demonstrate the effectiveness of CPICFs.

Counterfactuals improve model understanding and data augmentation.

Abstract

Counterfactual explanations for black-box models aim to pr ovide insight into an algorithmic decision to its recipient. For a binary classification problem an individual counterfactual details which features might be changed for the model to infer the opposite class. High-dimensional feature spaces that are typical of machine learning classification models admit many possible counterfactual examples to a decision, and so it is important to identify additional criteria to select the most useful counterfactuals. In this paper, we explore the idea that the counterfactuals should be maximally informative when considering the knowledge of a specific individual about the underlying classifier. To quantify this information gain we explicitly model the knowledge of the individual, and assess the uncertainty of predictions which the individual makes by the width of a conformal prediction interval. Regions of feature space where the prediction interval is wide correspond to areas where the confidence in decision making is low, and an additional counterfactual example might be more informative to an individual. To explore and evaluate our individualised conformal prediction interval counterfactuals (CPICFs), first we present a synthetic data set on a hypercube which allows us to fully visualise the decision boundary, conformal intervals via three different methods, and resultant CPICFs. Second, in this synthetic data set we explore the impact of a single CPICF on the knowledge of an individual locally around the original query. Finally, in both our synthetic data set and a complex real world dataset with a combination of continuous and discrete variables, we measure the utility of these counterfactuals via data augmentation, testing the performance on a held out set.