Why is one metric not enough to judge a binary classification model's performance?
F-score
Why is one metric not enough to judge a binary classification model's performance?
Imagine you're sorting emails into 'spam' or 'not spam.' You want to catch most spam but avoid mislabeling important emails as spam.
If you focus only on catching spam (high recall), you might miss important emails (low precision). If you only avoid mislabeling (high precision), you might let spam through (low recall). The F1 score balances these two concerns.
Example
You correctly label 80% of spam and 90% of important emails. Without the F1 score, you'd just look at these numbers separately, missing the balance between catching spam and preserving important emails.
Remember this
The F1 score combines precision and recall into one metric, showing the balance between catching spam and preserving important emails.
Text adapted from Wikipedia, licensed under CC BY-SA 4.0.
Related concepts
Evaluation of binary classifiers
Why can't we just use one score for everything?
Binary classification
Why is it important to know if something is a disease or not?
to use F1 score: when classes are imbalanced and both FP and FN matter
Why might a sports team need a different score to judge their performance?
score matching does: learns the gradient of the log-density without normalizing
Ever wonder how we can compare apples and oranges fairly in studies?
to use AUC-ROC: comparing classifiers across all thresholds
Why can't we always trust a yes-or-no answer?
Boosting (machine learning)
Boosting reduces bias in ML models
Swipe through more Machine Learning concepts
Open Pocket Polymath