Overview
MAE provides a simple, intuitive measure of how far off a model's predictions are on average. Unlike Mean Squared Error (MSE), it does not penalize large errors more heavily than small ones.
Formula
MAE = (1/n) * Σ |Actual - Predicted|
Characteristics
- Robust to Outliers: Because it doesn't square the errors, a few extreme values won't distort the total score as much as they would in MSE.
- Interpretability: The result is in the same units as the target variable (e.g., if you're predicting house prices in dollars, the MAE is also in dollars).