Crime Rate Analysis Using Machine Learning Technique

Abstract

A method for methodically detecting crime, analyzing crime patterns, and anticipating crime trends is crime analysis. The information gleaned from machine learning is of great use to police officers and can be applied to a large number of crime datasets. This issue could be resolved by utilizing a Random Forest in security analysis and law enforcement. Since the Random Forest algorithm has been cited as the most effective machine learning algorithm for predicting crime data, this work investigated the construction of a prototype model for crime prediction using the Random Forest algorithm. In addition to displaying criminal offense areas within a region, our algorithm is able to identify and forecast locations with a high likelihood of occurrence. The experimental results show that the Random Forest was able to correctly identify the unknown category in the crime data by 0.82, which is good enough to trust the system for predicting future crimes. This method's results can be used to raise awareness of risky areas and assist law enforcement in predicting future crimes in a particular area within a given time frame. Due to the expanding use of computerized and informational systems, data analysts of crime may be able to assist police departments in speeding up the process of solving crimes in our society. The machine learning system is simple to set up and works with the spatial plot of crime and criminal activities to improve the performance of our police and other government agencies. The Bangladesh police can reduce crime and solve cases as quickly as possible by implementing this developed system.