Machine Learning - Decision Trees - Michael Fuchs

Due to their structure, decision trees are easy to understand, interpret and visualize. In doing so, a variable check or feature selection is implicitly performed. Both numerical and non-numerical data can be processed simultaneously relatively little effort on the part of the user for the data preparation requires. On the other hand, too complex trees can be created that do not generalize the data well. Small variations in the data can also make the trees unstable, creating a tree that does not solve the problem.