Crime Rate Prediction System -An Experiment with Denver Crime Dataset Using Machine Learning Technique

Abstract

In recent years the nation Nigeria has experienced an increasing rate of criminality in the six geopolitical zones. Different crimes ranging from kidnapping, herdsmen attack, banditry, killings and so on. These activities have generated fear in the minds of the citizens thereby disrupting individuals, communities and their economic activities. This has affected both foreign and local investors in investing in the state. The overall effect on the socio-economic growth of the nation is unbearable. This paper presents a supervised machine learning technique for crime prediction using the Random Forest classifier algorithm and visualisation on the Denver crime dataset. The Denver crime dataset was used in this research due to its completeness and the lack of comprehensive dataset in the Nigerian police department. The prediction classification applied in this piece of work was based on the most frequent crime type, hotspot and crime count. The finding shows that the year 2022 experienced more crime related issues and theft crime was observed to be the highest while District 3 and 6 were seen as crime hotspots.