Here, we will learn how to actually conduct CART and Random Forest-based Land Use / Land Cover (LULC)-Classifications using only the Google Earth Engine and its provided datasets. This includes preparing the imagery and boundary data, creating training samples, running the classification and eventually validating it.

The Goal is to produce a LULC-Classified Map of the Syrian region of Raqqa for June 2020, using only available data that can be accessed from within the Google Earth Engine.

As collecting training samples can take quite a long time, you can alternatively click here to access the complete example script with the samples already included.