Error Types
Error types used when comparing datasets
| Keyword | Description |
|---|---|
RMSE |
Root mean squared error: $$\Delta x = \sqrt{\sum_{i=1}^{N} (x_i - x_i^{ref})^2 \frac{1}{N}}$$ |
MAAPE |
Mean arctangent average percentage error: $$\Delta x = 100 \sum_{i=1}^{N} \arctan{\left|\frac{x_i - x_i^{ref}}{x_i}\right|}$$ |
MAPE |
Mean average percentage error: $$\Delta x = 100 \sum_{i=1}^{N} \left|\frac{x_i - x_i^{ref}}{x_i}\right|$$ |
Percent |
Basic percentage error: $$\Delta x = 100 \frac{\sum_{i=1}^{N} \left|x_i - x_i^{ref}\right|}{\sum_{i=1}^{N} \left|y_i\right|}$$ |
RFactor |
EPSR-style r-factor, equivalent to the mean squared error: $$\Delta x = \sum_{i=1}^{N} (x_i - x_i^{ref})^2 \frac{1}{N}$$ |
Euclidean |
Euclidean distance: $$\Delta x = \frac{\sum_{i=1}^{N} \left|x_i - x_i^{ref}\right|}{\sum_{i=1}^{N} \left|x_i^2\right|}$$ |