How would a classification tree model be used for prediction? Presume that you have a person for whom the target variable has not yet been observed.

How would you use a tree model to predict a class label for this person?

1. Calculate the entropy.
2. Average the purity in each leaf node.
3. Find the leaf node with the same attributes as the individual and use the majority class in that leaf as the class label prediction.
4. Simply run the algorithm.