6.034 Artificial Intelligence - Recitations, fall 2004 online slides on learning

Next: Four Cases to Consider Previous: Decision Tree Learning

Top-Down Induction of Decision Trees

  1. At each point, decide which attribute to use as next test in the tree

  2. Attribute splits data based on answer to question

    Each answer forms a separate node in decision tree

    Each node is the root of an entire sub-decision tree problem, possibly with fewer examples and one fewer attribute than its parent