Evaluation Metrics
RMSE - Root Mean Square Error
RMSLE - Root Mean Squared Log Error
There is a considerable difference. RMSLE is a more recent measure with two advantages: (1) It is more robust to outliers and (2) it is a measure of relative errors, which is useful if you do not care about the differences in absolute value. On the other hand, RMSLE is designed to penalize under-estimation of the actual variable (relative to over-estimation) whereas RMSE is symmetric on either side. If you are looking for a more neutral measure regarding over- vs. under-estimation, RMSE remain useful.
For more, see https://medium.com/analytics-vidhya/root-mean-square-log-error-rmse-vs-rmlse-935c6cc1802a
The difference between RMSE and RMSLE dues to the nature of the data, simply, RMSE is used when the data tends to be linear, and RMSLE is used in some special cases when the data tends to be non-linear.
RMSE is the square root of MSE. In case of unbiased estimator, RMSE is just the square root of variance, which is actually Standard Deviation. In case of RMSLE, you take the log of the predictions and actual values. So basically, what changes is the variance that you are measuring. https://www.kaggle.com/c/ashrae-energy-prediction/discussion/113064
The value of the RMSE explodes in magnitude as soon as it encounters an outlier. In contrast, even on the introduction of the outlier, the RMSLE error is not affected much. From this small example, we can clearly infer that RMSLE is very robust when outliers come into play. https://medium.com/analytics-vidhya/root-mean-square-log-error-rmse-vs-rmlse-935c6cc1802a
RMSLE metric only considers the relative error between and the Predicted and the actual value and the scale of the error is not significant. On the other hand, RMSE value Increases in magnitude if the scale of error increases. https://medium.com/analytics-vidhya/root-mean-square-log-error-rmse-vs-rmlse-935c6cc1802a
- Badges
- Science topic
- Similar topics
- Control Systems Engineering
- Filtering