A Study on Recidivism Risk Prediction Based on Crime-Specific Human and Opportunity Costs

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Analysis shows a 45% recidivism rate in offenders, indicating opportunity costs impact prediction accuracy.

Recidivism rate among offenders stands at 45%, with age playing a crucial role in outcomes.
Logistic regression and decision tree methods yield predictive accuracies of 66.7% to 68.8%, surpassing baseline metrics.
The impact of prior convictions is nonlinear, with offenders having five or more priors showing significantly higher risks.
The approach integrates age and prior offenses into decision-making tools, providing transparent estimates without racial bias.

Lu et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark