So, you are still trying to predict the target variable, in this example predicting income, but with different more accurate model. Discuss findings, etc. see guidance. Don’t forget the rules.

Then, from our assignment guidance: Try to derive a more accurate model, or a model that is equally accurate but simpler (i.e., has fewer variables). Then you revise your variables, maybe fewer variables, simpler, etc. to attempt to get a better model. In this example you still want to predict income. Now you may have other variables, fewer variables to predict the outcome. What you are trying to predict is still the same, what changes is the variables, number of variables, or whatever you do to get a more accurate model, you are using to get the outcome.

From our assignment guidance: Be sure to specify what variable is being predicted and which variables are to be used to predict the outcomes. First you identify the target variable you want to predict. For example, you want to predict income. You may have variables such as education level, house price …. other variables to predict the outcome You run the first decision tree, make sure you discuss the rules

There was a question from student, and since my answer would benefit and help the class here is the clarification/feedback. This is regarding using a target variable for the following assignment guidance: Develop at least two different decision tree prediction models. Be sure to specify what variable is being predicted and which variables are to be used to predict the outcomes. Try to derive a more accurate model, or a model that is equally accurate but simpler (i.e., has fewer variables). Discuss your findings. Discuss the results and which model would be preferable. Which variables are the strongest predictors? What is the single best predictor variable? What combination of variables seems to give the best answer? Why?