Comparative Analysis of Machine Learning Models in Predicting Recidivism

Predicting whether a criminal is likely to commit another crime is an important research topic in the fields of criminology and sociology. This article explores various factors that affect the likelihood of recidivism among criminals, including age, gender, and social race. Through a comprehensive literature review and statistical analysis of the impact of different factors on the number of criminals, this study uses data science to investigate which factors affect the recidivism rate of crime, and proposes some key factors for predicting the likelihood of criminal recidivism. Finally, this article explores how to use these predictive factors to improve the prediction and intervention measures of crime recidivism rates, and establishes relevant models to predict the risk of crime recidivism, such as Random Forest model logistic regression model, decision tree model, and support vector machine model. The accuracy of these four models is compared and analyzed.