A Proof System with Causal Labels (Part I): checking Individual Fairness and Intersectionality

TL;DR

This paper introduces an extension to a typed natural deduction calculus to verify individual fairness and intersectionality in probabilistic classifiers using causal labels for conditional independence checks.

Contribution

It extends TNDPQ with causal labels to model fairness verification, enabling formal reasoning about individual fairness and intersectionality.

Findings

Formalizes fairness verification using causal labels

Enables checking of conditional independence in probabilistic models

Provides a logical framework for fairness analysis

Abstract

In this article we propose an extension to the typed natural deduction calculus TNDPQ to model verification of individual fairness and intersectionality in probabilistic classifiers. Their interpretation is obtained by formulating specific conditions for the application of the structural rule of Weakening. Such restrictions are given by causal labels used to check for conditional independence between protected and target variables.