Skip to contents

Measure to compare true observed response with predicted response in regression tasks.

Details

The pinball loss for quantile regression is defined as $$ \text{Average Pinball Loss} = \frac{1}{n} \sum_{i=1}^{n} w_{i} \begin{cases} q \cdot (t_i - r_i) & \text{if } t_i \geq r_i \\ (1 - q) \cdot (r_i - t_i) & \text{if } t_i < r_i \end{cases} $$ where \(q\) is the quantile and \(w_i\) are normalized sample weights.

Dictionary

This Measure can be instantiated via the dictionary mlr_measures or with the associated sugar function msr():

mlr_measures$get("regr.pinball")
msr("regr.pinball")

Meta Information

  • Task type: “regr”

  • Range: \((-\infty, \infty)\)

  • Minimize: TRUE

  • Average: macro

  • Required Prediction: “quantiles”

  • Required Packages: mlr3

Parameters

IdTypeDefaultRange
alphanumeric-\([0, 1]\)

See also

Other Measure: Measure, MeasureClassif, MeasureRegr, MeasureSimilarity, mlr_measures, mlr_measures_aic, mlr_measures_bic, mlr_measures_classif.costs, mlr_measures_debug_classif, mlr_measures_elapsed_time, mlr_measures_internal_valid_score, mlr_measures_oob_error, mlr_measures_regr.rsq, mlr_measures_selected_features

Super classes

mlr3::Measure -> mlr3::MeasureRegr -> MeasurePinball

Methods

Inherited methods


Method new()

Creates a new instance of this R6 class.

Usage

MeasureRegrPinball$new(alpha = 0.5)

Arguments

alpha

numeric(1)
The quantile to compute the pinball loss. Must be one of the quantiles that the Learner was trained on.


Method clone()

The objects of this class are cloneable with this method.

Usage

MeasureRegrPinball$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.