🤖 AI Summary
This work proposes an interpretable risk scoring method that directly optimizes net benefit—the primary objective in clinical decision-making—rather than conventional accuracy metrics. By formulating the problem as a sparse integer linear program, the approach simultaneously optimizes net benefit across multiple decision thresholds and yields transparent scoring rules with integer coefficients. Integrating techniques from interpretable machine learning and model calibration, the method achieves significantly higher net benefit on multiple public and real-world clinical datasets while maintaining strong discriminative performance and calibration. Furthermore, the study establishes theoretical connections between net benefit and traditional evaluation metrics, offering new insights into the relationship between clinical utility and standard predictive performance measures.
📝 Abstract
Risk scoring systems are widely used in high-stakes domains to assist decision-making. However, existing approaches often focus on optimizing predictive accuracy or likelihood-based criteria, which may not align with the main goal of maximizing utility. In this paper, we propose a novel risk scoring system that directly optimizes net benefit over a range of decision thresholds. The model is formulated as a sparse integer linear programming problem which enables the construction of a transparent scoring system with integer coefficients, and hence, facilitates interpretation and practical application. We also establish fundamental relationships among net benefit, discrimination, and calibration. Our analysis proves that optimizing net benefit also guarantees conventional performance measures. We thoroughly evaluated our method on multiple public datasets as well as on a real-world clinical dataset. This computational study demonstrated that our interpretable method can effectively achieve high net benefit while maintaining competitive discrimination and calibration performance.