An ML engineer needs to use an ML model to predict the price of apartments in a specific location. Which metric should the ML engineer use to evaluate the model's performance?

Question

Options

• AAccuracy
• BArea Under the ROC Curve (AUC)
• CF1 score
• DMean absolute error (MAE)

• A
  11% (6)
• B
  6% (3)
• C
  4% (2)
• D
  79% (42)

Explanation

Mean Absolute Error (MAE) is the right choice because apartment price prediction is a regression problem (predicting a continuous numeric value), and MAE directly measures the average magnitude of prediction errors in the same units as the target (e.g., dollars), making it intuitive and interpretable.

• A (Accuracy) is for classification problems - it measures the percentage of correct categorical predictions, which has no meaning when the target is a continuous number like price.
• B (AUC-ROC) evaluates a binary classifier's ability to distinguish between two classes (e.g., spam vs. not spam) - again, irrelevant to price prediction.
• C (F1 Score) balances precision and recall for classification tasks, especially with imbalanced classes - still a classification metric, not applicable to regression.

Memory tip: If the output is a category, use classification metrics (Accuracy, F1, AUC). If the output is a number on a scale, use regression metrics (MAE, MSE, RMSE). "Price" is always a number - think MAE.

No community discussion yet for this question.