Explain what precision and recall are.
ref:Precision, Recall & Accuracy. Accuracy is not the only model metric… | by Shivangi Sareen | Medium
I use following definitions:
- Recall tells you how much of the 1’s you can find.
- Precision tells you how much junk there is in your predicted 1’s.
Knowing this, it is straightforward to calculate these metrics:
- Recall: correctly predicted 1’s / all 1’s = TP / (TP + FN)
- Precision: correctly predicted 1’s / all predicted 1’s = TP / (TP+FP)