When preparing for an interview, I got the following question (on this medium link),
Given a medical test produce a 1% False Positive rate (FPR), and the population True Positive Rate (TPR) is 5%, what is the likelihood of a True Positive given by the test?
Can I answer this question without the initial distribution between actual Positive and Negative? If I assume they are equally likely, then my answer will be $\frac{5}{5+95+1+99}= \frac{5}{200}=0.025$. Is my thinking correct?
Because, from what I understand, the confusion matrix should be like below, right?
Preidiction P Preidiction N
actual P | 0.05 x p(actual Positive) | 0.95 x p(actual Positive) |
actual N | 0.01 x p(actual Negative) | 0.99 x p(actual Negative) |
