Measure of average magnitude‐weighted directional accuracy in forecast tasks.
Details
$$
\mathrm{MDV} = \frac{1}{n-1}
\sum_{i=2}^n \lvert y_i - y_{i-1}\rvert \times
\begin{cases}
+1, & \text{if }\mathrm{sign}(y_i - y_{i-1})
= \mathrm{sign}(\hat y_i - y_{i-1}),\\
-1, & \text{otherwise.}
\end{cases}
$$
where n is the number of observations.
Dictionary
This mlr3::Measure can be instantiated via the dictionary mlr3::mlr_measures or with the associated sugar function mlr3::msr():
Task type
Forecast measures are registered with task_type = "regr" so they compose with the standard regression
measures (e.g. mlr3::mlr_measures_regr.rmse) on the mlr3::PredictionRegr that forecast learners produce.
List them via the key prefix, not the task type, as the latter returns nothing:
Meta Information
Task type: “regr”
Range: \((-\infty, \infty)\)
Minimize: FALSE
Average: macro
Required Prediction: “response”
Required Packages: mlr3, mlr3forecast
References
Blaskowitz O, Herwartz H (2011). “On economic evaluation of directional forecasts.” International Journal of Forecasting, 27(4), 1058–1065.
See also
Chapter in the mlr3book: https://mlr3book.mlr-org.com/chapters/chapter2/data_and_basic_modeling.html#sec-eval
Package mlr3measures for the scoring functions.
as.data.table(mlr_measures)for a table of available Measures in the running session (depending on the loaded packages).Extension packages for additional task types:
mlr3proba for probabilistic supervised regression and survival analysis.
mlr3cluster for unsupervised clustering.
Other Measure:
mlr_measures_fcst.acf1,
mlr_measures_fcst.coverage,
mlr_measures_fcst.mase,
mlr_measures_fcst.mda,
mlr_measures_fcst.mdpv,
mlr_measures_fcst.mpe,
mlr_measures_fcst.msis,
mlr_measures_fcst.pinball,
mlr_measures_fcst.rmsse,
mlr_measures_fcst.wape,
mlr_measures_fcst.winkler
Super classes
mlr3::Measure -> mlr3::MeasureRegr -> MeasureMDV