A Proof System with Causal Labels (Part II): checking Counterfactual Fairness

TL;DR

This paper extends a typed natural deduction calculus to verify counterfactual fairness in probabilistic classifiers by modeling causal labels and ensuring robustness under their variation.

Contribution

It introduces a formal proof system that models and checks counterfactual fairness using structural conditions on causal labels.

Findings

Formalizes verification of counterfactual fairness

Ensures robustness of evaluation under causal label variations

Extends existing natural deduction calculus

Abstract

In this article we propose an extension to the typed natural deduction calculus TNDPQ to model verification of counterfactual fairness in probabilistic classifiers. This is obtained formulating specific structural conditions for causal labels and checking that evaluation is robust under their variation.