We obtain the Information Matrix test for finite Gaussian mixtures exploiting the EM principle, which implies the moments tested are the expectation given the data of the moments one would assess if one knew the latent component each observation belongs to. Simulations indicate analytical expressions for the asymptotic covariance matrix of those moments adjusted for sampling variability in parameter estimators provide reliable finite sample sizes and good power against various alternatives, especially combined with the parametric bootstrap. We apply the test to the univariate distribution of per capita income across countries and its joint distribution with per capita CO2 emissions.

The information matrix test for Gaussian mixtures / Amengual, D., Fiorentini, G., Sentana, E.. - In: JOURNAL OF ECONOMETRICS. - ISSN 0304-4076. - ELETTRONICO. - 255:(2026), pp. 106234.1-106234.34. [10.1016/j.jeconom.2026.106234]

The information matrix test for Gaussian mixtures

Amengual, Dante;Fiorentini, Gabriele;
2026

Abstract

We obtain the Information Matrix test for finite Gaussian mixtures exploiting the EM principle, which implies the moments tested are the expectation given the data of the moments one would assess if one knew the latent component each observation belongs to. Simulations indicate analytical expressions for the asymptotic covariance matrix of those moments adjusted for sampling variability in parameter estimators provide reliable finite sample sizes and good power against various alternatives, especially combined with the parametric bootstrap. We apply the test to the univariate distribution of per capita income across countries and its joint distribution with per capita CO2 emissions.
2026
255
1
34
Goal 3: Good health and well-being
Amengual, Dante; Fiorentini, Gabriele; Sentana, Enrique
File in questo prodotto:
File Dimensione Formato  
The information matrix test for Gaussian mixtures.pdf

Accesso chiuso

Tipologia: Pdf editoriale (Version of record)
Licenza: Solo lettura
Dimensione 5.01 MB
Formato Adobe PDF
5.01 MB Adobe PDF   Richiedi una copia all'autore

I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificatore per citare o creare un link a questa risorsa: https://hdl.handle.net/2158/1499111
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact