Concolic Testing on Individual Fairness of Neural Network Models
This work addresses algorithmic fairness in critical domains by providing a rigorous method for testing and verifying fairness in pre-trained DNNs, representing an incremental advance in formal fairness evaluation.
The paper tackles the problem of evaluating individual fairness in deep neural networks by introducing PyFair, a formal framework that adapts concolic testing to generate fairness-specific path constraints. Results show it effectively detects discriminatory instances and verifies fairness on 25 benchmark models, though scalability issues arise with complex models.
This paper introduces PyFair, a formal framework for evaluating and verifying individual fairness of Deep Neural Networks (DNNs). By adapting the concolic testing tool PyCT, we generate fairness-specific path constraints to systematically explore DNN behaviors. Our key innovation is a dual network architecture that enables comprehensive fairness assessments and provides completeness guarantees for certain network types. We evaluate PyFair on 25 benchmark models, including those enhanced by existing bias mitigation techniques. Results demonstrate PyFair's efficacy in detecting discriminatory instances and verifying fairness, while also revealing scalability challenges for complex models. This work advances algorithmic fairness in critical domains by offering a rigorous, systematic method for fairness testing and verification of pre-trained DNNs.