Interpretable Classification Models for Recidivism Prediction

TL;DR

This paper explores creating accurate, transparent, and interpretable recidivism prediction models, demonstrating that SLIM models outperform traditional interpretability-focused methods while matching the accuracy of black-box models across the ROC curve.

Contribution

It introduces the use of SLIM (Supersparse Linear Integer Models) to produce interpretable models that are as accurate as black-box models across the full ROC curve.

Findings

SLIM models achieve high accuracy and interpretability.

Traditional interpretability methods like CART and C5.0 cannot match accuracy.

Many machine learning methods produce similar accuracy along the ROC curve.

Abstract

We investigate a long-debated question, which is how to create predictive models of recidivism that are sufficiently accurate, transparent, and interpretable to use for decision-making. This question is complicated as these models are used to support different decisions, from sentencing, to determining release on probation, to allocating preventative social services. Each use case might have an objective other than classification accuracy, such as a desired true positive rate (TPR) or false positive rate (FPR). Each (TPR, FPR) pair is a point on the receiver operator characteristic (ROC) curve. We use popular machine learning methods to create models along the full ROC curve on a wide range of recidivism prediction problems. We show that many methods (SVM, Ridge Regression) produce equally accurate models along the full ROC curve. However, methods that designed for interpretability (CART, C5.0) cannot be tuned to produce models that are accurate and/or interpretable. To handle this shortcoming, we use a new method known as SLIM (Supersparse Linear Integer Models) to produce accurate, transparent, and interpretable models along the full ROC curve. These models can be used for decision-making for many different use cases, since they are just as accurate as the most powerful black-box machine learning models, but completely transparent, and highly interpretable.