Much of the data that we use for exploratory analysis is missing data. One way to handle the missing data is to impute it. We will use related data to impute crime locations.

What if we could determine the type of crime, forecast when a type of crime would happen again in a certain location or at a time of day, or what crimes are most predictable, or what features are most predictive of crimes? Maybe crime fighting could be improved, but this isn’t the first time people tried to address these issues. Simply googling forecast crime will render many interesting results.

(more…)