Precision

#machine-learning #interview ##evaluation

PREciSIon looks at only on the Predicted Side
Better Precision means less false positive
In "Spam Detection", Precision is more important than Recall
- As if some "Not Spam" mail is Falsely predicted Positive then it will skip the inbox

Precision

Tells us out of all predicted cases how much is correct for class A

P r e c i s i o n_{A} = \frac{T P_{A}}{T P_{A} + F P_{A}}

Related Notes