Broken
Artificial Intelligence
Your Classifier Is Broken, But It Is Still Useful
Once you run a binary classifier over a population you get an estimate of the proportion of true positives in that population. That is referred to as the prevalence.An ROC curve like this one...
Recent posts
