Some measures of model accuracy like mean absolute error (MAE), mean absolute percentage error (MAPE), symmetric mean absolute percentage error (SMAPE), mean squared error (MSE) and root mean squared error (RMSE).
MAE(x, ...)
# S3 method for default
MAE(x, ref, na.rm = FALSE, ...)
# S3 method for lm
MAE(x, ...)MAPE(x, ...)
# S3 method for default
MAPE(x, ref, na.rm = FALSE, ...)
# S3 method for lm
MAPE(x, ...)
SMAPE(x, ...)
# S3 method for default
SMAPE(x, ref, na.rm = FALSE, ...)
# S3 method for lm
SMAPE(x, ...)
MSE(x, ...)
# S3 method for default
MSE(x, ref, na.rm = FALSE, ...)
# S3 method for lm
MSE(x, ...)
RMSE(x, ...)
# S3 method for default
RMSE(x, ref, na.rm = FALSE, ...)
# S3 method for lm
RMSE(x, ...)
NMAE(x, ref, train.y)
NMSE(x, ref, train.y)
the specific numeric value
the predicted values of a model or a model-object itself.
the observed true values.
the observed true values in a train dataset.
a logical value indicating whether or not missing values should be removed. Defaults to FALSE.
further arguments
Andri Signorell <andri@signorell.net>
The function will remove NA values first (if requested).
MAE calculates the mean absolute error:
$$\frac{1}{n} \cdot \sum_{i=1}^{n}\left | ref_{i}-x_{i} \right |$$
MAPE calculates the mean absolute percentage error:
$$\frac{1}{n} \cdot \sum_{i=1}^{n}\left | \frac{ref_{i}-x_{i}}{ref_{i}} \right |$$
SMAPE calculates the symmetric mean absolute percentage error:
$$\frac{1}{n} \cdot \sum_{i=1}^{n}\frac{2 \cdot \left | ref_{i}-x_{i} \right |}{\left | ref_{i} \right | + \left | x_{i} \right |}$$
MSE calculates mean squared error:
$$\frac{1}{n} \cdot \sum_{i=1}^{n}\left ( ref_{i}-x_{i} \right )^2$$
RMSE calculates the root mean squared error:
$$\sqrt{\frac{1}{n} \cdot \sum_{i=1}^{n}\left ( ref_{i}-x_{i} \right )^2}$$
Armstrong, J. S. (1985) Long-range Forecasting: From Crystal Ball to Computer, 2nd. ed. Wiley. ISBN 978-0-471-82260-8
https://en.wikipedia.org/wiki/Symmetric_mean_absolute_percentage_error
Torgo, L. (2010) Data Mining with R: Learning with Case Studies, Chapman and Hall/CRC Press
r.lm <- lm(Fertility ~ ., data=swiss)
MAE(r.lm)
# the same as:
MAE(predict(r.lm), swiss$Fertility)
MAPE(r.lm)
MSE(r.lm)
RMSE(r.lm)
Run the code above in your browser using DataLab