Up to this point we, the humans in the story, have been making these decision trees by hand. We have not even discussed how we write a program to allow the robot to use trees to make decisions. Wouldn’t it be a lot nicer if the computer was doing all the hard part of making the tree, deciding the branches, and labeling the nodes instead of us? That is exactly what we will do in this section.
Let’s consider the problem of classifying toys. We may want to come up with a more efficient robot, which sorts toys in some manner instead of just dumping them in a box. In an ideal world, out of a population of 20 toys, we would have some characteristic that divided the group evenly in half – 10 and 10. ...
