Relative Squared Error

Details

The Relative Squared Error is defined as $$\frac{\sum _{i=1}^n{(t_i-r_i)}^2}{\sum _{i=1}^n{(t_i-\overline{t})}^2},$$ where $\overline{t}=\sum _{i=1}^n{t}_i$.

Can be interpreted as squared error of the predictions relative to a naive model predicting the mean.

This measure is undefined for constant $t$.

Note

The score function calls mlr3measures::rse() from package mlr3measures.

If the measure is undefined for the input, NaN is returned. This can be customized by setting the field na_value.

Dictionary

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

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

Parameters

Empty ParamSet

Meta Information

• Type: "regr"

• Range: $[0,\mathrm{\infty })$

• Minimize: TRUE

• Required prediction: response

See also