🔬 Mean Absolute Error (MAE)

Mean Absolute Error (MAE)#

The average absolute difference between predictions and actuals.

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

What it is#

Mean Absolute Error is the average absolute gap between prediction and truth — the L1 error:

\[\text{MAE} = \frac{1}{N}\sum_{i=1}^{N} |y_i - \hat{y}_i|.\]

It reports the typical error in the same units as the target, with no squaring.

How it behaves#

Because it takes absolute values rather than squares, MAE weights all errors linearly and is far more robust to outliers than MSE / RMSE — one huge miss doesn’t dominate. The forecast that minimizes MAE is the median of the target (for RMSE it is the mean).

When to use it#

MAE is the right choice when you want an interpretable, outlier-resistant measure of typical error and don’t need to punish large mistakes extra hard. Its main limits: it is scale-dependent (not comparable across series — use MASE for that) and, being point-only, it can’t score probabilistic forecasts.

Theme: Model Evaluation & Uncertainty · All terminology

Hint

Mind map — connected ideas

Pinball Loss (a.k.a. Quantile Loss) · Continuous Ranked Probability Score (CRPS) · Root Mean Squared Error (RMSE) · Mean Squared Error (MSE) · MASE (Mean Absolute Scaled Error) · Mean Absolute Percentage Error (MAPE)

Hint

More in Model Evaluation & Uncertainty

Average Absolute Error (AAE) · Baseline Heuristics · Bootstrap · Bootstrap Confidence Intervals (CIs) · Coverage · Cramér’s V · DeLong’s Test · KS Statistic (Kolmogorov–Smirnov Statistic) · Likelihood Ratio (LR) · Mann–Whitney U Test (also called the Wilcoxon rank-sum test) · MASE (Mean Absolute Scaled Error) · Mean Absolute Percentage Error (MAPE) · Mean Squared Error (MSE) · Relative accuracy