Data & Analytics
MAE (Mean Absolute Error)
Also known as:
Mean Absolute Error
L1 Loss
Absolute Deviation
Updated: 2/12/2026
The average of absolute differences between prediction and reality – robust to outliers.
Quick Summary
MAE = mean absolute error – more robust than MSE/RMSE with outliers.
Explanation
MAE = 1/n × Σ|y_pred - y_true|. Easy to interpret and robust.
Marketing Relevance
MAE is preferred when outliers exist and all errors should be weighted equally.
Common Pitfalls
Gradient not differentiable at y=0. Less sensitive to large errors.
Origin & History
MAE is one of the oldest statistical metrics, already in use in the 18th century.
Comparisons & Differences
MAE (Mean Absolute Error) vs. MSE / RMSE
MSE/RMSE squares errors; MAE treats all errors linearly equal.