The empirical likelihood goodness-of-fit test for regression model

Lixing ZHU*, Yong Song Qin, Wang Li Xu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)

Abstract

Goodness-of-fit test for regression modes has received much attention in literature. In this paper, empirical likelihood (EL) goodness-of-fit tests for regression models including classical parametric and autoregressive (AR) time series models are proposed. Unlike the existing locally smoothing and globally smoothing methodologies, the new method has the advantage that the tests are self-scale invariant and that the asymptotic null distribution is chi-squared. Simulations are carried out to illustrate the methodology.

Original languageEnglish
Pages (from-to)829-840
Number of pages12
JournalScience in China, Series A: Mathematics, Physics, Astronomy
Volume50
Issue number6
DOIs
Publication statusPublished - Jun 2007

Scopus Subject Areas

  • Mathematics(all)

User-Defined Keywords

  • AR time series models
  • Asymptotic normality
  • Empirical likelihood
  • Goodness-of-fit
  • Regression model

Fingerprint

Dive into the research topics of 'The empirical likelihood goodness-of-fit test for regression model'. Together they form a unique fingerprint.

Cite this