2.4 Posterior probability model
The posterior probability model is defined as: P(is fake | has !) and P(is real | has !)
and this can be calculated using Bayes’ Rule:
posterior=prior×likelihoodnormalizing constant
Fake | Real | |
---|---|---|
prior | 40.0% | 60.0% |
likelihood | 26.7% | 2.2% |
posterior | 88.9% | 11.1% |
Shortcut to calculating the normalizing constant:
normalising constant=sum(prior×likelihood)