ToolJoltTools

Regression Metrics Calculator (MAE, MSE, RMSE, R²)

Paste predicted vs actual values and get MAE, MSE, RMSE, MAPE and R² — every regression error metric at once.

MAE
MSE
RMSE
MAPE (%)

RMSE penalizes large errors more than MAE (it squares them), so RMSE ≫ MAE signals a few big misses. MAPE is scale-free but explodes near zero actuals. R² < 0 means your model is worse than predicting the mean.

Formula

MAE = mean|ŷ−y| · MSE = mean(ŷ−y)² · RMSE = √MSE · MAPE = mean|ŷ−y|/|y| · R² = 1 − SS_res/SS_tot
References: Hyndman & Koehler (2006), Another look at measures of forecast accuracy; Willmott & Matsuura (2005), Advantages of MAE over RMSE

About Regression Metrics Calculator (MAE, MSE, RMSE, R²)

Every regression model needs error metrics, but each tells a different story and reporting only one hides problems. This calculator takes your predicted and actual value pairs and computes the full set: MAE (average absolute miss, robust and interpretable), MSE/RMSE (outlier-sensitive, in the target's units), MAPE (percentage error, scale-free but unstable near zero) and R² (variance explained). Seeing them together is diagnostic — a large RMSE-to-MAE ratio reveals outliers, a negative R² reveals a model worse than the mean, and MAPE flags whether percentage thinking even applies.

How to use Regression Metrics Calculator (MAE, MSE, RMSE, R²)

  1. 1Enter your values into Regression Metrics Calculator (MAE, MSE, RMSE, R²) — sensible, domain-typical defaults are pre-filled so you see a real result immediately.
  2. 2The result recomputes live using the formula shown on the page; there is no button to press.
  3. 3Adjust any input to compare scenarios, then read the worked example to see the substituted numbers.

Why use Regression Metrics Calculator (MAE, MSE, RMSE, R²)?

  • Computes Regression Metrics instantly in your browser — no sign-up, no upload, no server round-trip.
  • 100% free and unlimited, with the exact formula shown: MAE = mean|ŷ−y|.
  • Runs entirely client-side, so every value you enter stays private on your device.
  • Live recompute as you type, with a worked example and authoritative references for trust.

Frequently asked questions

RMSE or MAE — which should I report?+

MAE if all errors are equally bad and you want an interpretable 'average miss'. RMSE if large errors are disproportionately costly (it squares them, so big misses dominate). Report both — their ratio is informative: RMSE ≈ MAE means uniform errors, RMSE ≫ MAE means a few large outliers are present.

What does a negative R² mean?+

Your model predicts worse than just always guessing the mean of the actuals. R² = 1 is perfect, 0 means you've matched the mean baseline, and negative means you're below it — usually a sign of a serious bug, a badly mismatched model, or evaluating on data very different from training.

Why is MAPE risky?+

It divides by actual values, so it blows up toward infinity as actuals approach zero and is undefined at zero. It also asymmetrically punishes over-prediction. For data spanning zero or with small values, prefer MAE, RMSE, or a symmetric/scaled alternative like sMAPE or MASE.

Are these metrics affected by the scale of my data?+

MAE, MSE and RMSE are in your target's units (or units²), so they're not comparable across different problems. MAPE and R² are scale-free and comparable. To compare models on the same data, any of them works; to compare across datasets, use MAPE or R² (with their caveats).

Related tools

Related ML & AI tools

Sponsored