Artificial Intelligence for Robotics

Random forests

I really wanted to add this section on random forest classifiers, just because the name sounds so darn cool. There has even been talk of extreme random forests. While I may have been accused of stretching metaphors to the breaking point, this time the name may have inspired the software. We learned how to make decision trees, and we have learned that they have some weak points. It is best if the data really belongs in distinct and differentiated groups. It is not very tolerant of noise in the data. And it really gets unwieldy if you want to scale it up – you can imagine how big a graph would get with 200 classes rather than the six or seven we were dealing with.

If you wanted to take advantage of the simplicity and utility ...