Clusters of effects curves in quantile regression models

Risultato della ricerca: Articlepeer review

4 Citazioni (Scopus)

Abstract

In this paper, we propose a new method for nding similarity ofefects based on quantile regression models. Clustering of effects curves (cec)techniques are applied to quantile regression coefficients, which are one-to-one functions of the order of the quantile. We adopt the quantile regressioncoefficients modeling (qrcm) framework to describe the functional form of thecoefficient functions by means of parametric models. The proposed methodcan be utilized to cluster the effect of covariates with a univariate responsevariable, or to cluster a multivariate outcome. We report simulation results,comparing our approach with the existing techniques. The idea of combiningcec with qrcm permits simplifying computation and interpretation of theresults, and may improve the ability to identify clusters.We illustrate a varietyof applications, highlighting the advantages and the usefulness of the describedmethod.
Lingua originaleEnglish
pagine (da-a)551-569
Numero di pagine19
RivistaComputational Statistics
Volume34
Stato di pubblicazionePublished - 2018

All Science Journal Classification (ASJC) codes

  • ???subjectarea.asjc.2600.2613???
  • ???subjectarea.asjc.1800.1804???
  • ???subjectarea.asjc.2600.2605???

Fingerprint

Entra nei temi di ricerca di 'Clusters of effects curves in quantile regression models'. Insieme formano una fingerprint unica.

Cita questo