Recall (machine learning)

A metric for classification models that answers the following question:¹

When ground truth was the positive class, what percentage of predictions did the model correctly identify as the positive class?

Formula

Recall is the proportion of all actual positives that were classified correctly as positives:

$$\frac{\text{correctly classified actual positives}}{\text{all actual positives}}$$

which means:

$$\text{Recall} = \frac{TP}{TP + FN}$$

where:

• True positive means the model correctly predicted the positive class.
• False negative means that the model mistakenly predicted the negative class.

Precision improves as false positives decrease, while recall improves when false negatives decrease. They often show an inverse relationship, where improving one of them worsens the other. (See also F1 score)

When (not) to use

Use when false negative are more expensive than false positives.²

Recall is particularly useful for determining the predictive power of classification models in which the positive class is rare. Recall is a much more useful metric for class-imbalanced datasets than accuracy.¹ But the number of actual positives is very, very low, say 1-2 examples in total, recall is less meaningful and less useful as a metric.²

See also

• False positive rate (a.k.a. probability of false alarm, where the recall is known as called probability of detection)
• Accuracy
• Precision
• ROC curve: Recall (a.k.a. True positive rate) is the y-axis in an ROC curve.

Footnotes

1. developers.google.com/machine-learning/glossary#recall ↩ ↩²

2. ML crash course - Classification ↩ ↩²