Dynamic Fairness
Dynamic fairness addresses the evolving nature of fairness in algorithmic decision-making, aiming to move beyond static assessments to account for temporal changes and long-term consequences. Current research focuses on developing methods to quantify and mitigate unfairness arising from dynamic system interactions, including the use of causal inference, reinforcement learning, and continual learning frameworks to analyze and improve fairness over time in various applications like federated learning and human-robot interaction. This research is crucial for building more equitable and trustworthy AI systems, impacting fields ranging from healthcare and finance to social robotics.
Papers
September 11, 2024
April 16, 2024
April 10, 2024
November 2, 2023
May 8, 2023
February 17, 2023
January 21, 2023
June 1, 2022