We present a method using randomized classifiers to improve individual fairness in the strategical classification setting and an algorithm finding the optimal threshold distribution.

This work studies the problem of designing classifiers that can incentivize strategic agents to exert effort in desirable features even when they have incomplete information about the classifier or the causal graph.

We extend the study of strategic classification to non-linear classifiers and study how behavior in this regime affects the shape of classifiers and expressivity of model classes.

We analyze how collective strategic behavior influences predictive models, introducing level-$k$ reasoning and showing how user coordination affects equilibrium outcomes.