So as suggested by Wang et al. [20], the prediction error of the τ-th conditional quantile was assessed based on the average of the value ρτ(Y−q^τ(Y|X=x)) over all observations, where ρτ(r) = τr − rI(r < 0) is called the quantile loss function, for 0 < τ < 1[10,20].
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}