Strategic Conformal Prediction
This addresses the challenge of reliable uncertainty estimation in strategic settings, which is crucial for deploying ML models in real-world applications where predictions influence behavior, representing a novel extension beyond standard conformal prediction.
The paper tackles the problem of uncertainty quantification for machine learning models when predictions alter the environment due to strategic agents, proposing Strategic Conformal Prediction to provide robust uncertainty quantification with theoretical guarantees and experimental validation showing effectiveness against arbitrary strategic alterations.
When a machine learning model is deployed, its predictions can alter its environment, as better informed agents strategize to suit their own interests. With such alterations in mind, existing approaches to uncertainty quantification break. In this work we propose a new framework, Strategic Conformal Prediction, which is capable of robust uncertainty quantification in such a setting. Strategic Conformal Prediction is backed by a series of theoretical guarantees spanning marginal coverage, training-conditional coverage, tightness and robustness to misspecification that hold in a distribution-free manner. Experimental analysis further validates our method, showing its remarkable effectiveness in face of arbitrary strategic alterations, whereas other methods break.