To fully evaluate the effectiveness of a predictive model, you must examine both precision and recall. Unfortunately, precision and recall are often in tension. That is, improving precision typically reduces recall and vice versa. Therefore, prioritizing between precision and recall is important. To increase recall (sensitivity), we should minimize FN (false negative). To increase precision, we should minimize FP (false positive), though. Now, assume that a bank manager wants to develop a predictive model to decide for loan applications (accept or reject). The predictive model should predict who defaults ("default" or " not default"). If "default", the loan application will be rejected. If "not default", it will be accepted. However, as a manager, she should decide which error (FP or FN) should be minimized. What do you think? Which error do you think should be minimized? What is the consequence of such decision for the business of a bank