(13. Machine learning and UDJ [35 points) An important characteristic of a classification model is the falsenegative rate: e.g., if the model classies a patient as healthy when, in fact, the patient is sick. To evaluate this false negative rate, one needs to have enough patients who are sick, which might be challenging if the disease is rare. a) How many patients with actual disease we need to run through the model in order to obtain an 80% condence interval for the falsenegative rate (i.e., for the proportion of patients with the disease who are incorrectly classified as healthy) with a margin of error of no more than 5%? (In other words, the width of the interval should be no more than 10%, or 0.1.)