Short Definition
Recall measures how many actual positives a model successfully finds.
Intuition
Recall asks: out of everything we needed to catch, how much did we catch?
Technical Definition
Recall is true positives divided by true positives plus false negatives.
Example
In cancer screening, high recall means fewer true cases are missed.
Common Misunderstandings
Recall does not measure how many positive predictions were false alarms.
High recall can come at the cost of lower precision.