Math
Regression Metrics Calculator
Measure model error with MAE, MSE, RMSE, MAPE, and R-squared
Enter matching actual and predicted series to evaluate the model.
Compare Prediction Quality Quickly
Regression metrics help you measure how close predictions are to actual values. This calculator gives you the most common error metrics in one place so you can judge absolute error, squared error, percent error, and explained variance together.
How To Read The Metrics
Lower MAE, MSE, and RMSE values indicate better predictive accuracy. Lower MAPE means smaller average percentage errors. Higher R-squared means the model explains more of the variation in the actual series.
Useful For Model Comparison
Use these metrics when comparing forecast models, regression pipelines, baselines, or tuning runs. Looking at several metrics together is safer than relying on a single score.
Frequently Asked Questions
What is the difference between MAE and RMSE?
MAE averages absolute errors, so every miss contributes linearly. RMSE squares errors before averaging, so it penalizes larger misses more strongly and is often more sensitive to outliers.
Why can MAPE be misleading?
MAPE divides by the actual value. If actual values are zero or very close to zero, the percentage error can become unstable or undefined. This calculator skips zero actuals in the MAPE denominator.
What does R-squared mean?
R-squared measures how much of the variance in the actual values is explained by the predictions. Values closer to 1 indicate a better fit, while negative values can happen when predictions are worse than using the average.
When should I use mean error?
Mean error is useful for spotting systematic bias. A positive mean error means predictions tend to overshoot. A negative mean error means they tend to undershoot.