A recidivist is a criminal that was released from prison, but commits another crime. You are a warden at a maximum-security prison in Kingston, and you want to determine which prisoners will likely become recidivists. Luckily, you have a Queen's degree, so you are going to take a data-driven approach. You have collected some historical training data that include some basic metadata, and whether the prisoner ended up becoming a recidivist or not.

TASKS

1. [Math and Diagrams] Given the training data below, use the ID3 algorithm and entropy-based information gain to construct a decision tree by hand. Show all the steps and follow the algorithm closely. Show the resulting decision tree. Show the "summary statistics" at each leaf node.

 

id Good Behavior Age < 30 Drug Dependent Recidivist 1 False True False True 2 False False False False 3 False True False True 4 True False False False True False True True 6 True False False false