a lean and sustainable conscientious cannabusiness; target market? Black market.

Predicting Recidivism and Analysis of National Corrections Data 1991 - 2014

CMU 95791-Data Mining Group Project on Recidivism Forecasting

Inspired by the National Institute of Justice's Recidivism Forecasting Challenge. It aims to refine recidivism forecasts using both person- and place-based variables to aid community supervision and reintegration efforts.

Data and analysis for 'Machine Bias'

Predicting and characterizing recidivism in Colombia as part of a group project (Team 77) for DS4A certification.

Fairness-aware recidivism risk assessment - calibrated predictions, SHAP explanations, conformal uncertainty, human-in-the-loop overrides, and a full audit trail.

ML-based feature selection + conditional discrimination mitigation with unified evaluation and outputs

XAI analysis on the COMPAS dataset for recidivism

NIJ Recidivism Forecasting Challenge — deep-learning model + 2026 revised findings: re-offending concentrates in a persistent high-prior-history cohort. Reproducible analysis included.

A repository of various articles, abstracts, essays, and studies which analyze successful reentry strategies and report methods of avoiding recidivism.

Reoffending-risk prediction with a Neo4j knowledge graph and a Fairlearn audit on the COMPAS dataset. A small auditable architecture demonstration.

R code to aggregate and merge features from different government sources, create flat file, perform relationship, prediction, and equity measures

Machine learning analysis of the COMPAS recidivism dataset examining predictive accuracy, algorithmic fairness, and behavioral risk factors associated with two-year reoffending.

The Bureau of Justice Statistics (BJS) publishes information on crime, criminal offenders, victims of crime, and the operation of justice systems.