0

I was thinking if F1 score is usually correlated with support in classification problems. In theory, shouldn't the F1 score increase for a label if there is more support? Why does this not always happen?

Maria
  • 3
  • 1
  • 2
    Possible duplicate of [what does the numbers in the classification report of sklearn mean?](https://stats.stackexchange.com/questions/117654/what-does-the-numbers-in-the-classification-report-of-sklearn-mean) – xiawi Jul 02 '19 at 07:08
  • I do know what each of these mean...what I would like to know if it is common that support effects f1 score and whether there is a direct correlation – Maria Jul 02 '19 at 07:13
  • 1
    Having never heard the nomenclature "F1 score", for a moment I thought the question concerned the relationship between fan support of F1 teams (as in Formula 1) and their results at the end of the championship – Easymode44 Jul 02 '19 at 07:34
  • The support is the number of times each class appears in your *data*... Why would that necessarily correlate positively with a measure of performance of your *model*? – Frans Rodenburg Jul 10 '19 at 09:05

1 Answers1

0

The support only relates to the ground truth. It is the same whathever your result is. On the contrary, the F1-score is computed from your actual scores thus depends from your results.

Thus, there is no direct link between support and F1-score in theory.

You can have an indirect link if the model performs better on a particular class. In that case, the more sample in that class, the larger the support and the higher the F1-score. But it is really due to the fact that

  • the model performs well on that class thus the F1-score increases

  • the class is larger thus the support is larger

xiawi
  • 118
  • 5