Fair Contextual Multi-Armed Bandits: Theory and Experiments
This work addresses fairness in allocation decisions for AI systems interacting with users, such as virtual agents or robots, but it is incremental as it builds on existing bandit methods with fairness constraints.
The paper tackles the problem of ensuring fairness in AI systems that allocate tasks or resources among multiple users, by introducing a contextual multi-armed bandit algorithm with fairness constraints, and shows through theory and experiments that accounting for contexts improves fair decision-making, especially when user performance varies by context.
When an AI system interacts with multiple users, it frequently needs to make allocation decisions. For instance, a virtual agent decides whom to pay attention to in a group setting, or a factory robot selects a worker to deliver a part. Demonstrating fairness in decision making is essential for such systems to be broadly accepted. We introduce a Multi-Armed Bandit algorithm with fairness constraints, where fairness is defined as a minimum rate that a task or a resource is assigned to a user. The proposed algorithm uses contextual information about the users and the task and makes no assumptions on how the losses capturing the performance of different users are generated. We provide theoretical guarantees of performance and empirical results from simulation and an online user study. The results highlight the benefit of accounting for contexts in fair decision making, especially when users perform better at some contexts and worse at others.