What is the difference between "Random Forest" and "Forest-based Classification and Regression" in ArcGIS Pro?

Question

ArcGIS Pro (2.6.0) has tools to train Random Forest (named Random Trees in ArcGIS) and Support Vector Machine.  Afterwards, the tool named "Classify Raster" contains the algorithms to apply your trained algorithm to imagery.  You can choose between Random Trees and SVM.  There is a unique geoprocessing tool named "Forest-based Classification and Regression" that is supposedly an adaptation of Random Trees.  What is the difference?

Qbert · Answer

There's not a tremendous difference... Forest-based Classification & Regression is an application of Random Forest to build and use predictive models with vector data (though it can use rasters as explanatory variables). It could not use the term "Random Forest" because the term is trademarked.
The other tool you referenced is actually called Random Trees, and it's used for applying the Random Forest algorithm for raster analysis.

What is the difference between "Random Forest" and "Forest-based Classification and Regression" in ArcGIS Pro?

One Answer

Add your own answers!

Ask a Question