Crime Prediction Utilizing ARIMA Model

Abstract

With the development of society and the progress of human beings, the issue of crime has attracted more attention. This research uses one of the most common time series models named the ARIMA model as the main algorithm to predict the crime data by month in San Francisco from 2003 to 2015, obtained from Kaggle. This research is about to study all categories of crime and specific crime cases as well. The focus is on the number of cases that happened every month. By applying the ARIMA model, it can be tested whether the trend of happened crime cases fits the model and predict the future crime tendency. The verification and prediction of crime data are studied with the help of the ARIMA model. And the goal of studying this dataset is to improve the safety and security of human beings. In this research, the ARIMA model simulates the previous crime trend effectively and forecasts the future crime trend with high accuracy. This is helpful to improve the safety level and reduce the crime rates in San Francisco.