A new study explores how Explainable Machine Learning can predict and analyze homicide clearance rates across the U.S. Using data from the Murder Accountability Project, researchers tested nine algorithms, with XGBoost emerging as the most accurate. The study also employed SHAP (a tool for AI explainability) to reveal key factors affecting case resolution, including victim demographics, weapon type, and crime circumstances. The findings reveal substantial variability in homicide clearance rates at the state level, emphasizing the need for localized investigative strategies.

Read the full study here: https://arxiv.org/pdf/2203.04768

#MachineLearning #AI #Criminology #LawEnforcement #DataScience #HomicideAnalysis #XGBoost #ExplainableAI #PublicSafety