In this chapter, we chose a pertinent problem that had both analytics and geospatial components and tried to apply a very robust ML technique known as random forest to it. Before building the model, we had to handle the date component, the spatial component of data, as well as the categorical and continuous variables. We were able to achieve a good score in our first pass and build a world-class model with a few lines of code and a little bit of spatial data processing.
In the next chapter, we will discuss more accurate real-world distance metrics and perform other spatial computations, such as intersection, to make the model better.
