A first look at the F1 score

The classification metric you will meet properly in Model Evaluation.

Regression metrics measure distance; classification needs different ones. The F1 score is the standard metric when classes are imbalanced, and you will study it properly in Model Evaluation.

It combines precision (of the items you flagged, how many were right?) and recall (of the true positives, how many did you find?) with a harmonic mean: $F_{1} = \frac{2 P R}{P + R}$ .

The harmonic mean is pulled towards the smaller value, so a model can't get a high F1 by being great at one and terrible at the other.

Formula to remember

Least squares loss

J (w) = i = 1 \sum n (y_{i} - w^{⊤} x_{i})^{2}

Sum of squared residuals.

Quick check · +5 XP

R² = 0 means the model…

Your notes

Saved in this browser as you type.