Predicting Vancouver Crime with Decision Trees

Predictive policing is not a new idea. In fact, it’s one that comes with understandable controversy! Early predictive policing was partly inspired by the advanced analytics in E-commerce domains. Wal-Mart and Amazon were early adopters of using user statistics to anticipate emerging trends, and consumer behaviour. Instead of predicting consumer behaviour from user data, predictive policing forecasts criminal behaviour from crime data. People generally have privacy and digital ethics concerns with creating individual profiles for how likely people are to engage in illegal behaviour. One use of predictive policing is not predicting where a crime will occur, but who will be involved. In this case individual profiles must be constructed. On the positive side, this technique allows police to identify potential terrorists before they cause mass harm. At the same time, however, creating individual profiles is a breach of privacy to many and can have dire consequences. For example, imagine you are flagged as a high-risk “terrorist” when you aren’t. <a href="https://medium.com/@alex.blackwell314/predicting-vancouver-crime-with-decision-trees-9f7795240303">Learn More</a>