Which statement correctly describes the F1 Score?

Get ready for the GARP Risk and AI Exam with flashcards and multiple choice questions. Each question comes with hints and explanations. Prepare for success!

Multiple Choice

Which statement correctly describes the F1 Score?

The F1 score captures a balance between precision and recall by taking their harmonic mean. Precision asks how often predicted positives are truly positive, while recall asks how many of the actual positives you managed to find. By using the harmonic mean, the F1 score becomes sensitive to both sides: if either precision or recall is low, the F1 score drops significantly, even if the other value is high. This makes F1 a useful single metric when you care about both false positives and false negatives, especially with class imbalance. The exact formula is F1 = 2 × (precision × recall) / (precision + recall). It’s not a measure of how well the model separates positive and negative classes (that’s more about separation metrics like ROC/AUC). It’s also not simply the sum of precision and recall, since the harmonic mean tends to pull the result toward the smaller of the two values. And it isn’t generally equal to accuracy, because accuracy depends on true negatives and true positives together, whereas F1 depends only on the positive predictions through precision and recall.

The F1 score captures a balance between precision and recall by taking their harmonic mean. Precision asks how often predicted positives are truly positive, while recall asks how many of the actual positives you managed to find. By using the harmonic mean, the F1 score becomes sensitive to both sides: if either precision or recall is low, the F1 score drops significantly, even if the other value is high. This makes F1 a useful single metric when you care about both false positives and false negatives, especially with class imbalance.

The exact formula is F1 = 2 × (precision × recall) / (precision + recall). It’s not a measure of how well the model separates positive and negative classes (that’s more about separation metrics like ROC/AUC). It’s also not simply the sum of precision and recall, since the harmonic mean tends to pull the result toward the smaller of the two values. And it isn’t generally equal to accuracy, because accuracy depends on true negatives and true positives together, whereas F1 depends only on the positive predictions through precision and recall.

More practice questions