Geodesic
On the symmetrized chi-square tests in autoregression with outliers in data
Teoriâ veroâtnostej i ee primeneniâ, Tome 68 (2023) no. 4, pp. 691-704 Cet article a éte moissonné depuis la source Math-Net.Ru

Voir la notice de l'article

A linear stationary model $\mathrm{AR}(p)$ with unknown expectation, coefficients, and the distribution function of innovations $G(x)$ is considered. Autoregression observations contain gross errors (outliers, contaminations). The distribution of contaminations $\Pi$ is unknown, their intensity is $\gamma n^{-1/2}$ with unknown $\gamma$, and $n$ is the number of observations. The main problem here (among others) is to test the hypothesis on the normality of innovations $\boldsymbol H_{\Phi}\colon G (x)\in \{\Phi(x/\theta),\,\theta>0\}$, where $\Phi(x)$ is the distribution function of the normal law $\boldsymbol N(0,1)$. In this setting, the previously constructed tests for autoregression with zero expectation do not apply. As an alternative, we propose special symmetrized chi-square type tests. Under the hypothesis and $\gamma=0$, their asymptotic distribution is free. We study the asymptotic power under local alternatives in the form of the mixture $G(x)=A_{n,\Phi}(x):=(1-n^{-1/2})\Phi(x/\theta_0)+n^{-1/2}H(x)$, where $H(x)$ is a distribution function, and $\theta_0^2$ is the unknown variance of the innovations under $\boldsymbol H_{\Phi}$. The asymptotic qualitative robustness of the tests is established in terms of equicontinuity of the family of limit powers (as functions of $\gamma$, $\Pi,$ and $H(x)$) relative to $\gamma$ at the point $\gamma=0$.
Mots-clés : autoregression
Keywords: outliers, residuals, empirical distribution function, chi-square test, local alternatives, robustness. 
@article{TVP_2023_68_4_a1,
     author = {M. V. Boldin},
     title = {On the symmetrized chi-square tests in autoregression with outliers in data},
     journal = {Teori\^a vero\^atnostej i ee primeneni\^a},
     pages = {691--704},
     year = {2023},
     volume = {68},
     number = {4},
     language = {ru},
     url = {http://geodesic.mathdoc.fr/item/TVP_2023_68_4_a1/}
}
TY  - JOUR
AU  - M. V. Boldin
TI  - On the symmetrized chi-square tests in autoregression with outliers in data
JO  - Teoriâ veroâtnostej i ee primeneniâ
PY  - 2023
SP  - 691
EP  - 704
VL  - 68
IS  - 4
UR  - http://geodesic.mathdoc.fr/item/TVP_2023_68_4_a1/
LA  - ru
ID  - TVP_2023_68_4_a1
ER  - 
%0 Journal Article
%A M. V. Boldin
%T On the symmetrized chi-square tests in autoregression with outliers in data
%J Teoriâ veroâtnostej i ee primeneniâ
%D 2023
%P 691-704
%V 68
%N 4
%U http://geodesic.mathdoc.fr/item/TVP_2023_68_4_a1/
%G ru
%F TVP_2023_68_4_a1
M. V. Boldin. On the symmetrized chi-square tests in autoregression with outliers in data. Teoriâ veroâtnostej i ee primeneniâ, Tome 68 (2023) no. 4, pp. 691-704. http://geodesic.mathdoc.fr/item/TVP_2023_68_4_a1/

[1] M. V. Boldin, “On the Pearson's chi-square test for normality of autoregression with outliers”, Theory Probab. Appl., 65:1 (2020), 102–110 | DOI | DOI | MR | Zbl

[2] T. W. Anderson, Jr., The statistical analysis of time series, Wiley Ser. Probab. Stat., John Wiley Sons, Inc., New York–London–Sydney, 1971, xiv+704 pp. | DOI | MR | Zbl

[3] P. J. Brockwell, R. A. Davis, Time series: theory and methods, Springer Ser. Statist., Springer-Verlag, New York, 1987, xiv+519 pp. | DOI | MR | Zbl

[4] R. D. Martin, V. J. Yohai, “Influence functionals for time series”, Ann. Statist., 14:3 (1986), 781–818 | DOI | MR | Zbl

[5] M. V. Boldin, M. N. Petriev, “On the empirical distribution function of residuals in autoregression with outliers and Pearson's chi-square type tests”, Math. Methods Statist., 27:4 (2018), 294–311 | DOI | MR | Zbl

[6] M. V. Boldin, On stochastic expansions of the empirical distribution function of residuals in autoregression schemes, 2021, 14 pp., arXiv: 2108.05903