Detecting cancer tends to follow this model as you account for false negatives. Although not ideal, it is better to be safe and label a non-threatening tumor as threatening rather than labeling a threatening tumor as non-threatening. 

Example of Low Precision and High Recall

The model describes the class well, but the data is unbalanced and the model picks up other classes.

University of Michigan - Ann Arbor

There are four combinations of precision and recall:
(1) High Precision and High Recall
(2) High Precision and Low Recall
(3) Low Precision and High Recall
(4) Low Precision and Low Recall

Relation between Precision and Recall

The model describes the class extremely well.

High Precision and High Recall

The model has a hard time describing the class but when it does describe the given class, it is good at doing so.

High Precision and Low Recall

Low Precision and High Recall

Model can not describe the class well. Handles the data poorly. Low precision and low recall is the least ideal for describing data.

Learn Before

Related

Learn After