Technology Sharing

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

F1-score is an indicator to measure the performance of classification models. It is particularly suitable for processing extremely unbalanced data sets. The value range of F1-score is from 0 to 1. The larger the value, the better the performance.
Calculation formula:
F1-score is the harmonic mean of precision and recall.
$\ast \ast F1score=2\times \frac{precision\times Recall}{precision+Recall}\ast \ast$
So what are precision and recall?
Precision: The ratio of samples predicted by the model to samples that are actually positive.
Recall: The proportion of samples that are actually positive that are correctly predicted as positive by the model.
Calculated as follows:
$Precision=\frac{TP}{TP+FP}$
$Recall=\frac{TP}{TP+FN}$
Eg：

What is TP? FP? FN？
TP (True Positive): The model predicts that the model is positive and the model is actually positive. There are 2 such examples.
FP（False Positive）: The model predicts that the class is positive, but it is actually negative. There is 1 in the example.
FN (False Negative): The model predicts a negative class, but it is actually a positive class. There is 1 here.
$Precision=\frac{TP}{TP+FP}=\frac{2}{2+1}=0.67$
$Recall=\frac{TP}{TP+FN}=\frac{2}{2+1}=0.67$
$F1score=2\times \frac{Precision\times Recall}{Precision+Recall}=2\times \frac{0.67\times 0.67}{0.67+0.67}=0.67$

What questions do recall and precision answer?？
Recall: is theActually positiveIn the sample,Predicted to be positiveof!
So the answer is: among all the positive samples, how many are correctly identified by the model?
Accuracy: Yes in allPredicted as positiveIn the sample,Actually, it is also positiveThe ratio.
So the answer is: among all the samples predicted to be positive, how many are actually positive?