This project seeks to perform an objective evaluation of relative safety in the city Boston using police incident data. By utilizing the publicly available incident data, Boston’s crime can be represented by a surface in three space. Overlaying the heatmap of the surface onto a map of Boston will provide an intuitive representation of the relative danger in parts of the city.
The data used for this project is the 2012-2015 police incident dataset from the city of Boston’s Open Government initiative. This dataset contains 268,057 incidents. Incidents are defined by anytime a police office is dispatched to an area. The incident data includes precise location of the incident (15 significant figures of longitude and latitude). The incident data also includes the type of crime, whether or not a weapon was involved, as well as the date and time the officer arrived on the scene. Incidents are grouped into regions about 1/6th of a square mile in area. Relative danger is then calculated by applying MATLAB's Gaussian smooth (with block size of 10 and standard deviation of 0.65).
and Professor Dwight Horan for facilitating this project.
Code will made available once the first offical version is complete.