Classification Under Strategic Self-Selection
This addresses a novel setting in strategic classification for machine learning systems where user participation is influenced by predictions, which is incremental as it builds on existing work but focuses on self-selection rather than feature modifications.
The paper tackles the problem of strategic user behavior in classification, where users decide whether to participate based on the classifier's predictions, and proposes a differentiable framework for learning under such self-selective behavior, demonstrating its utility through experiments on real data and simulations.
When users stand to gain from certain predictions, they are prone to act strategically to obtain favorable predictive outcomes. Whereas most works on strategic classification consider user actions that manifest as feature modifications, we study a novel setting in which users decide -- in response to the learned classifier -- whether to at all participate (or not). For learning approaches of increasing strategic awareness, we study the effects of self-selection on learning, and the implications of learning on the composition of the self-selected population. We then propose a differentiable framework for learning under self-selective behavior, which can be optimized effectively. We conclude with experiments on real data and simulated behavior that both complement our analysis and demonstrate the utility of our approach.