Machine learning: Bayesian decision theory

The PPT:

We skipped the AIDS detection problem in the Bayesian decision theory PPT. Now, use the Bayes theorem to confirm the answer given on the PPT file. For this problem, you need to distinguish two different conditions: False positive is a conditional probability P(reagent is negative | patient is infected). Same argument for false negative. When a patient is given a positive test result, it is actually P(patient is infected reagent is positive) Probability is important o It is known that the HIV test has only 0.1% of false positive and false negative, respectively. However, for a specific group of people, the prevalence of HIV positive rate is 0.01 %. If a person belongs to such a group and is found to be positive in the HIV test, find the probability that the person is really infected Answer: 9% (why? Think conditional probability) Same for mass testing of COVID-19