Identification of Criminal Activity Hotspots using Machine Learning to Aid in Effective Utilization of Police Patrolling in Cities with High Crime Rates

Criminal activity has always been a major deterrent in human progress and the constant presence of criminal activities stemming out of multifarious causes has been a hindrance for human sustainable living. The problem further aggravates when there is a dearth of police force to prevent crime. In countries like India where the police to population ratio is much less than the United Nations' Standard, the need of the hour is to efficiently utilize the existing force to prevent crimes. In this paper, we propose a solution that would facilitate effective distribution of police forces in a city among multiple districts based on the extent to which each district is prone to crime at a given hour, in a given day, for a given month. We have used the Chicago Crime Dataset in this work. The problem has been modelled as an imbalanced classification problem and supervised machine learning algorithms such as Logistic Regression, Naive Bayes, K Nearest Neighbours, Support Vector Machines, Decision Trees, Random Forests, Gradient Boosting Trees have been employed and their performances have been evaluated. In particular, the Gradient Boosting Tree has achieved the best performance in our case.