Precision, Recall & F1 Score

Precision = TP/TP+FP

Precision means of all the points, the model predicted to be positive, what percentage are actually positive.

Recall = TPR(True Positive Rate) TP/P

TPR = TP/TP+FN

Recall means of all the points that actually belonged to class 1, how many of them have been predicted to be positive.

Precision can lie between 0-1, similarily Recall can lie between 0-1.

For a good model, we want precision to be high and recall to be high.

F1 Score = 2* Precision*Recall/(Precision + Recall)

F1 Score is used in a lot of kaggle competitions.

Share

Share on twitter
Share on facebook
Share on linkedin
Share on whatsapp
Share on reddit
Share on telegram
Share on pinterest
Share on email
Share on facebook
Close Menu