Root Mean Squared Error: the square root of MSE, which puts the error back in the units of the target. Squaring first lets large misses dominate, so RMSE is always at least MAE and far more outlier-sensitive — report MAE too when a handful of extreme rows drives the number.
Card 1 of 259. Answer: Root Mean Squared Error: the square root of MSE, which puts the error back in the units of the target. Squaring first lets large misses dominate, so RMSE is always at least MAE and far more outlier-sensitive — report MAE too when a handful of extreme rows drives the number.
About · Quizzes · History of AI · To suggest new questions or report an error send me a dm.
RMSE, root mean squared error, is the square root of the average squared difference between predictions and actual values: RMSE = √((1/n) Σ (yᵢ − ŷᵢ)²). It is the most common error metric for regression and is in the same units as the target.
Example: errors of 2, −1 and 4 give squared errors of 4, 1 and 16, a mean of 7 and an RMSE of about 2.65. The mean absolute error (MAE) of the same errors is 2.33. RMSE is always at least as large as MAE, and the gap grows when a few errors are large, because squaring lets big misses dominate.
Use RMSE when large errors are disproportionately costly; use MAE when every unit of error costs the same or the data has outliers you don't want to dominate the metric. RMSE is the metric least-squares regression implicitly optimizes. It is scale-dependent, so it can only be compared across models predicting the same target.
