A sample of the MEPS data set to predict whether a patient will visit a hospital's emergency department (ED) at least once in a given year. To build this prediction model, we use the variable ED_Visit, which is 1 if the respondent recorded at least one ED visit in the survey year and 0 otherwise.

Question 1

Are we solving a supervised or unsupervised prediction problem? Are we solving a regression problem or a classification problem?

Question 2

Describe one example of a false positive prediction and one example of a false negative prediction for this prediction problem. For each type of prediction error, also describe the cost imposed by this type of error and identify who bears that cost. Would you weigh false positive predictions and false negative predictions equally? If not, which type of prediction error would you weigh more heavily and why?