Glossary · Phase 3 · Machine learning
RMSE
The square root of the mean squared error: an error in the target's own units that still punishes big misses more than small ones.
Try it now
Squared error punishes one big miss far more than several small ones. That's why a single outlier can drag a squared-error model.
Learn it properly
Loss functions · One number that says how wrong the model is. About 25 minutes.
Other terms from the same lesson
- Huber loss: A loss that treats small misses like squared error and big misses like absolute error, so a few wild values can't dominate training.
- loss: A single number saying how wrong a model's predictions are. Training tries to make it small.
- MAPE: The average size of the errors as a percentage of the actual values; easy to read, but undefined when an actual value is zero.
- mean absolute error: The average size of the misses, ignoring whether each was too high or too low. A miss twice as big costs twice as much.
- mean squared error: A loss: square each prediction's error and average them, so big mistakes count extra.
- quantile loss: A lopsided loss that charges more for misses in one direction, such as running out of stock costing more than overstocking.
Never trained an AI?
Teach a small one to write, right in your browser. It takes about 20 seconds and needs no account.