Machine Learning Prediction Models on the Tendency of Criminals to Recurrence of Criminal Activities

To enhance internal state security, it is crucial to understand the reasons behind recidivism, as it is the criminal, not the public threat, who causes the crime. With the digitization of law enforcement information operations, integrating blockchain technology with machine learning techniques proves practical for predicting and preventing criminal activity. The application of ML algorithms for identifying key factors influencing recidivism among prisoners has been addressed in this article. Specifically, by examining various individual factors of convicts’, decision tree framework was employed to predict the likelihood of repeat infringements. The number of actual and suspended consecutive charges are the primary determinants of recidivism. Decision Tree models were developed to classify prisoners as either susceptible or not susceptible to reoffending, providing valuable insights for criminal justice decision-making and improving inmate management within the correctional system.