/sf-crime

Machine learning to predict the category of crime incidents

Primary LanguageJupyter Notebook

SF Crime Classification

Using data from the SFPD Crime Incident Reporting system (https://www.kaggle.com/c/sf-crime) from 2003 through 2015, we built a model to predict what crime category an incident belongs to. The dataset primarily includes features relating to time and location. Three general categories of crime were considered: violent crime, deviant behavior, and theft (the most common type by far).