Implementing Decision Trees Just an if-then-else statement Choosing the features is more important For numeric features, thresholds are to be picked With increasing features including numerical ones, difficult to set up the structure by hand Decision Tree structure can be learned using machine learning over a training corpus Basic Idea Usually works top-down, by choosing a variable at each step that best splits the set of items. Popular algorithms: ID3, C4.5, CART Text Processing: Basics Week 1: Lecture 5 6 /26