Question 15
Dr. Burada is a cardiologist. He is trying to predict if the occurrence of a “Heart Attack” can be predicted using “Age”, “Resting bpm” and “Blood Cholesterol Level”. Accordingly, the data for 200 patients was collected and a logistic regression model was built (model coefficients are provided in Table-4). As Dr. Burada’s wife was running a sweets shop, he wanted to identify patients who were not prone to “Heart Attack” (so that he can sell them more oily snack items and get a name with the Home Minister). Hence, the target is “No Heart Attack=1” and “Heart Attack=0”. Table-5 gives the list of patients visiting the sweet shop. Dr.Burada needs to spend approximately 20 mins per Non-Heart Attack patient canvasing if he wants to convince them to buy snacks. Given this information, answer the given subquestions.
If the threshold is decreased to 0.5, which of the following is/are true? (choose all that is applicable)
Precision for predicting Y=1 will increase
Precision for predicting Y=1 will decrease
Recall for predicting Y=1 will increase
Recall for predicting Y=1 will decrease
Precision for predicting Y=0 will increase
Precision for predicting Y=0 will decrease
Recall for predicting Y=0 will increase
Recall for predicting Y=0 will decrease
Accuracy of the model will decrease
Accuracy of the model will increase
None of these