
Root mean square error, a metric measuring the average magnitude of prediction errors.Root mean square error (RMSE) is a standard measure of how far predictions are from actual values. It is calculated by taking the differences between predicted and observed values (the errors), squaring them to remove negative signs, averaging those squares, and then taking the square root. The result is in the same units as the original data, making it intuitive: an RMSE of 3 meters means typical prediction errors are around 3 meters.
The concept has roots in least-squares methods dating back to Gauss and Legendre in the early 19th century, though the specific term 'RMSE' became common with the rise of statistical forecasting in the 20th century.
For each data point, compute the error (predicted minus actual). Square each error, sum them all, divide by the number of points to get the mean squared error, then take the square root. Squaring gives more weight to large errors, so RMSE is sensitive to outliers. It is always non-negative, and a value of zero means perfect predictions.
Used everywhere from weather forecasting to machine learning to evaluate model accuracy—for example, comparing different climate models' predictions of temperature against observed records.
RMSE is always larger than or equal to the mean absolute error (MAE), because squaring emphasizes big mistakes, so the difference between them tells you how much your errors are driven by outliers.













