MLJun 25, 2013
Supersparse Linear Integer Models for Predictive Scoring Systems
arXiv:1306.5860v120 citations
AI Analysis
This provides a tool for building interpretable predictive models, but it appears incremental as it builds on existing linear integer modeling approaches.
The authors tackled the problem of creating interpretable binary classification scoring systems by introducing Supersparse Linear Integer Models (SLIM), which they showed to be accurate, sparse, and interpretable through theoretical risk bounds and experimental results.
We introduce Supersparse Linear Integer Models (SLIM) as a tool to create scoring systems for binary classification. We derive theoretical bounds on the true risk of SLIM scoring systems, and present experimental results to show that SLIM scoring systems are accurate, sparse, and interpretable classification models.