621. Prior Shift in Bayesian Classifiers
medium
A spam filter uses Bayes' theorem to classify emails. After deployment, the volume of spam increases from 20% to 50% of all emails. How does this affect the classifier's output?
A spam filter uses Bayes' theorem to classify emails. After deployment, the volume of spam increases from 20% to 50% of all emails. How does this affect the classifier's output?