Machine Learning Approach in Crime Records Evaluation

Due to the increase in the number of criminal records, research is going on with the help of data analytics to track the criminal nature and behavior for better understanding and to secure our society from the criminal. There are many ways to analyze the criminal records, but few models focus on prediction. In this paper, we have discussed the models used for predicting criminal behavior. We used four types of the model here for prediction the crime, and those models includes the LWL algorithm, linear regression, naive Bayes, and decision trees. LWL algorithm performance shows that it can be used for general reliably crime prediction with the frequency of antisocial behavior.