Research on crime risk prediction model based on machine learning

Abstract: Crime prediction is an important method for public security departments to conduct crime early warning and investigation. According to the multidimensional characteristics of criminals, machine learning classification algorithm is used to establish a prediction model of whether a criminal commits a crime again, so as to quantitatively evaluate the possibility of the criminal commits a crime again, and provide a feasible scheme reference for preventing and reducing the recurrence of crime. Analysis of urban crime data in Boston, using Naive Bayes, Random Forest and XGboost algorithms to build a crime type prediction model, we can get the urban crime type tendency, help the public security department to predict the future crime, take effective measures to reduce the urban crime rate.