Machine Learning

Metrics

Measure what matters for your problem and business cost.

Classification

  • Accuracy: OK when classes are balanced.
  • Precision/Recall/F1: best when false positives/negatives have different cost.
  • ROC-AUC / PR-AUC: evaluate ranking quality across thresholds.

Regression

  • MAE: average absolute error (robust, easy to interpret).
  • RMSE: penalizes large errors more.
  • R²: relative fit (can mislead if used alone).
from sklearn.metrics import mean_absolute_error, mean_squared_error
mae = mean_absolute_error(y_test, pred)
rmse = mean_squared_error(y_test, pred, squared=False)