8. A support vector machine has been built to predict whether a patient is at risk of cardiovascular disease. In the dataset used to train the model there are two target levels—high risk (the positive level, +1) or low risk

(the negative level, −1)—and three descriptive features—AGE, BMI, and BLOOD PRESSURE. The support vectors in trained the model are shown in the table below (all descriptive feature values have been standardised).

In the model the value of w0 is −0.0216, and the values of the α
parameters are 〈1.6811, 0.2384, 0.2055, 1.7139〉. What predictions would this model make for the following query instances?

Transcribed Image Text:

AGE -0.4549 BMI BLOOD PRESSURE RISK 0.0095 0.2203 low risk -0.2843 -0.5253 0.3668 low risk 0.3729 0.0904 -1.0836 high risk 0.558 0.2217 0.2115 high risk