evaluation
Recall (ML)
The proportion of actual positive cases that the model correctly identifies, measuring how many real positives the model catches. High recall means few false negatives.
In practice
If there are 200 spam emails and the model catches 180 of them, its recall is 90%.