A Comparative Analysis of Artificial Intelligence Techniques in Forecasting Violent Crime Rate

Abstract

The increase in the occurrence of violent crimes is a major concern in all countries around the globe. Various approaches of crime analyses have been implemented in reducing the number of violent crimes and among them is crime forecasting. Crime forecasting is an effective solution as it assists law enforcement agencies in planning efficient crime prevention measures. It has been observed recently that the application of artificial intelligence (AI) techniques in crime forecasting and analysis is favoured by researchers. Motivated by this development, this study aims to conduct a comparative analysis on the forecasting performance of three artificial intelligence (AI) techniques, namely artificial neural network (ANN), support vector regression (SVR), and gradient tree boosting (GTB) in forecasting the rates of four types of crimes in the United States (US). The forecasting performance of each AI technique was compared in terms of quantitative error measurement. From the results obtained, GTB showed the highest forecast accuracy compared to ANN and SVR as the observed error measurements were the smallest.