Crime Prediction using Machine Learning

Abstract

Crime is one of the world’s most serious problems, affecting people’s daily lives, whether they are travelling home from work or going on a trip. The situation in Vancouver is the same. This paper focuses on forecasting crime-prone areas (using the XGBOOST algorithm) based on data currently available for that area. Along with prediction, we built it so that the crime hotspots for any date, month, year, and time combination can display on a Google map, making the work of the governing agency easier and more efficient, helping to reduce Vancouver’s crime rate.