Learning Optimized Risk Scores on Large-Scale Datasets

12 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Developing an optimal short-form of the PTSD Checklist for DSM-5 (PCL-5) (2019, methodology): “The first integer‐scored method used Risk‐calibrated Supersparse Linear Integer Model (RiskSLIM; Ustun & Rudin, 2017), which is a machine learning algorithm to efficiently find the best‐fitting logistic regression model that has small integer weights and obeys custom constraints.”
Simple Rules for Complex Decisions (2017, methodology): “Many prior rule construction methods oer great exibility and performance [35], but in turn require considerable computational expertise to carry out.”
A Provable Algorithm for Learning Interpretable Scoring Systems (2018, background): “The state-of-the-art methods [27, 7, 16, 30, 28] are reported to be accurate, but an obvious drawback is that their output, the learned scores, apply to real-valued data (if the input data were real).”
Interpretable Optimal Stopping (2018, unknown): “A stream of research has considered interpretable scoring rules for classification and risk prediction; recent examples include Ustun and Rudin (2015, 2016) and Zeng et al. (2017).”
Optimized Scoring Systems: Toward Trust in Machine Learning for Healthcare and Criminal Justice (2018, methodology): “In our studies of recidivism (62, 52, 54), we used the largest publicly available”
Optimized Risk Scores (2017, methodology): “We set C0 to a small value (10−8) to recover the best model under these constraints [33].”