Prediction of Crime

Abstract

Crime has always been a controversial topic amongst society. The prediction of criminal events could help law enforcement units to be more alert in the areas, reducing or preventing the events from ever happening. Even in cases when the crime is committed, law enforcement would be already deployed in the area reducing the chances of the criminal getting away. Historical data was retrieved from the databases of the government of Puerto Rico, where only 3 years of data were available. A fit model regression considering categorical and ordinal variables resulted in an equation that validated our data with a degree of error. Although, the equation estimates the location on the next crime, its accuracy could be improved with more data and analysis of the factors that influence crime. The program is not intended to determine the location and time with 100% accuracy, but to give a good estimate of these for a crime. Key Terms  Crime Prediction, Crime Statistics, Government Open Data, Regression Models.