We consider new classes of estimators and test statistics for models satisfying linear constraints with unknown parameter. These procedures are based on minimization of divergences through duality techniques. We prove that, for various divergences, the new approach provides robust estimation and test procedures, unlike the empirical likelihood method. We give general results using the influence function approach, which we exemplify in detail in the case of the Cressie–Read divergences. It is found that the Hellinger distance is one of the divergences that leads to robust procedures.
Nous considérons de nouvelles classes dʼestimateurs et de procédures de test pour des modèles satisfaisant des contraintes linéaires à paramètre inconnu. Ces procédures sont basées sur la minimisation des divergences grâce à des techniques de dualité. Nous prouvons que, pour de nombreuses divergences, la nouvelle approche fournit des estimateurs et des tests robustes, contrairement à la méthode de vraisemblance empirique. Nous donnons des résultats généraux en utilisant lʼapproche par fonction dʼinfluence, que nous illustrons en détail dans le cas des divergences de Cressie–Read. On remarque que la distance de Hellinger conduit à des procédures robustes.
Accepted:
Published online:
Aida Toma 1, 2
@article{CRMATH_2013__351_7-8_311_0, author = {Aida Toma}, title = {Robustness of dual divergence estimators for models satisfying linear constraints}, journal = {Comptes Rendus. Math\'ematique}, pages = {311--316}, publisher = {Elsevier}, volume = {351}, number = {7-8}, year = {2013}, doi = {10.1016/j.crma.2013.02.005}, language = {en}, }
Aida Toma. Robustness of dual divergence estimators for models satisfying linear constraints. Comptes Rendus. Mathématique, Volume 351 (2013) no. 7-8, pp. 311-316. doi : 10.1016/j.crma.2013.02.005. https://comptes-rendus.academie-sciences.fr/mathematique/articles/10.1016/j.crma.2013.02.005/
[1] New estimates and tests of independence in semiparametric copula models, Kybernetika, Volume 46 (2010), pp. 178-201
[2] Minimization of ϕ-divergences on sets of signed measures, Stud. Sci. Math. Hung., Volume 43 (2006), pp. 403-442
[3] Parametric estimation and tests through divergences and the duality technique, J. Multivar. Anal., Volume 100 (2009), pp. 16-31
[4] Divergences and duality for estimation and test under moment condition models, J. Stat. Plan. Infer., Volume 142 (2012), pp. 2554-2573
[5] An estimation method for Neyman chi-square divergence with application to test of hypothesis, J. Multivar. Anal., Volume 97 (2006), pp. 1409-1436
[6] Decomposable pseudodistances and applications in statistical estimation, J. Stat. Plan. Infer., Volume 142 (2012), pp. 2574-2585
[7] Multinomial goodness-of-fit tests, J. R. Stat. Soc. Ser. B, Volume 46 (1984), pp. 440-464
[8] Weighted empirical likelihood estimates and their robustness properties, Comput. Stat. Data An., Volume 51 (2007), pp. 5130-5141
[9] Robust Statistics: the Approach Based on Influence Functions, Wiley, New York, 1986
[10] Dual representation of ϕ-divergences and applications, C. R. Acad. Sci. Paris, Ser. I, Volume 336 (2003), pp. 857-862
[11] Empirical Likelihood, Chapman & Hall CRC, 2001
[12] Empirical likelihood and general estimating equations, Ann. Statist., Volume 22 (1994), pp. 300-325
[13] Dual divergence estimators and tests: Robustness results, J. Multivar. Anal., Volume 102 (2011), pp. 20-36
[14] Robust tests based on dual divergence estimators and saddlepoint approximations, J. Multivar. Anal., Volume 101 (2010), pp. 1143-1155
Cited by Sources:
Comments - Politique