Long-Term Fair Decision Making through Deep Generative Models
This work addresses fairness for demographic groups in sequential systems, but it appears incremental as it builds on existing causal and generative methods.
The paper tackles long-term fairness in sequential decision-making by proposing a three-phase learning framework that uses deep generative models to train decision models on high-fidelity data, with empirical evaluation showing efficacy on synthetic and semi-synthetic datasets.
This paper studies long-term fair machine learning which aims to mitigate group disparity over the long term in sequential decision-making systems. To define long-term fairness, we leverage the temporal causal graph and use the 1-Wasserstein distance between the interventional distributions of different demographic groups at a sufficiently large time step as the quantitative metric. Then, we propose a three-phase learning framework where the decision model is trained on high-fidelity data generated by a deep generative model. We formulate the optimization problem as a performative risk minimization and adopt the repeated gradient descent algorithm for learning. The empirical evaluation shows the efficacy of the proposed method using both synthetic and semi-synthetic datasets.