Mean Absolute Error in Python: Definition and Example

The mean_absolute_error in Python is a metric from sklearn.metrics that measures the average absolute difference between predicted and actual values. It shows how far predictions are from true values on average, with lower values meaning better accuracy.

How It Works

Mean Absolute Error (MAE) works by calculating the average of the absolute differences between predicted values and actual values. Imagine you are guessing the weight of apples, and you want to know how far off your guesses are on average. MAE tells you this average error without caring if the guess was too high or too low.

It’s like checking how much you missed by each time, then finding the average of those misses. This helps you understand the typical size of your errors in simple terms, making it easy to compare different models or predictions.

Example

This example shows how to calculate mean absolute error using sklearn.metrics.mean_absolute_error. We compare some true values with predicted values and get the average absolute error.

python

from sklearn.metrics import mean_absolute_error

y_true = [3.0, -0.5, 2.0, 7.0]
y_pred = [2.5, 0.0, 2.0, 8.0]

mae = mean_absolute_error(y_true, y_pred)
print(f"Mean Absolute Error: {mae}")

Output

Mean Absolute Error: 0.5

When to Use

Use mean absolute error when you want a simple and clear measure of prediction accuracy that treats all errors equally. It’s useful in cases like predicting house prices, temperatures, or any continuous values where you want to know the average size of your mistakes.

MAE is especially helpful when you want to avoid large errors having too much influence, unlike squared error metrics. It gives an easy-to-understand number that matches the units of your data, making it practical for real-world decisions.

Key Points

• MAE measures average absolute difference between predictions and true values.
• It is easy to interpret because it uses the same units as the data.
• MAE treats all errors equally without squaring them.
• Lower MAE means better prediction accuracy.
• Commonly used in regression problems for model evaluation.

Key Takeaways

Mean absolute error shows average size of prediction errors in original units.

Use sklearn's mean_absolute_error function to easily calculate MAE in Python.

MAE treats all errors equally, making it simple and intuitive.

Lower MAE values indicate more accurate predictions.

Ideal for regression tasks where understanding average error magnitude matters.