Avoiding Overfitting of Decision Trees

Abstract

This chapter begins by examining techniques for dealing with clashes (i.e. inconsistent instances) in a training set. This leads to a discussion of methods for avoiding or reducing overfitting of a decision tree to training data. Overfitting arises when a decision tree is excessively dependent on irrelevant features of the training data with the result that its predictive power for unseen instances is reduced.

Two approaches to avoiding overfitting are distinguished: pre-pruning (generating a tree with fewer branches than would otherwise be the case) and post-pruning (generating a tree in full and then removing parts of it). Results are given for pre-pruning using either a size or a maximum depth cutoff. A method of post-pruning a decision tree based on comparing the static and backed-up estimated error rates at each node is also described.