Characterization of Large Deviation Probabilities for Regenerative Sequences
Informatics and Automation, Branching Processes and Related Topics, Tome 316 (2022), pp. 47-63

Voir la notice de l'article provenant de la source Math-Net.Ru

Local theorems are considered for additive functionals of regenerative sequences, which are sequences of random vectors $\{S_n\}_{n\ge 0}$ of special form. Two cases of renewal are considered: proper and terminating renewal. Under the assumption that all renewal cycles satisfy the Cramér condition, in the case of proper renewal, A. A. Borovkov, A. A. Mogulskii and E. I. Prokopenko, as well as A. V. Shklyaev and G. A. Bakay, obtained exact asymptotics for large deviation probabilities $\mathbf P(S_n=x)\sim {D(x/n)}n^{-d/2}\exp (-L(x/n)n)$, $n\to \infty $, which are uniform with respect to $x/n=x(n)/n\in \mathbb R^d$ in compact sets, with certain functions $D$ and $L$. In the case of terminating renewal, similar results were obtained by Bakay; moreover, one more deviation zone was distinguished in which the result has the form $\mathbf P(S_n=x) \sim {D_0(x/n)}{n^{-(d-1)/2}}\exp (-L_0(x/n)n)$, $n\to \infty $, with certain functions $D_0$ and $L_0$. This relation holds uniformly with respect to $x/n=x(n)/n\in \mathbb R^d$ in compact sets. In the present paper, an alternative method is found for calculating the functions appearing in the asymptotics, and equivalent conditions are obtained for the theorems.
Keywords: local theorems, large deviations, random sequences with renewal, terminating renewal.
@article{TRSPY_2022_316_a4,
     author = {G. A. Bakay},
     title = {Characterization of {Large} {Deviation} {Probabilities} for {Regenerative} {Sequences}},
     journal = {Informatics and Automation},
     pages = {47--63},
     publisher = {mathdoc},
     volume = {316},
     year = {2022},
     language = {ru},
     url = {http://geodesic.mathdoc.fr/item/TRSPY_2022_316_a4/}
}
TY  - JOUR
AU  - G. A. Bakay
TI  - Characterization of Large Deviation Probabilities for Regenerative Sequences
JO  - Informatics and Automation
PY  - 2022
SP  - 47
EP  - 63
VL  - 316
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/TRSPY_2022_316_a4/
LA  - ru
ID  - TRSPY_2022_316_a4
ER  - 
%0 Journal Article
%A G. A. Bakay
%T Characterization of Large Deviation Probabilities for Regenerative Sequences
%J Informatics and Automation
%D 2022
%P 47-63
%V 316
%I mathdoc
%U http://geodesic.mathdoc.fr/item/TRSPY_2022_316_a4/
%G ru
%F TRSPY_2022_316_a4
G. A. Bakay. Characterization of Large Deviation Probabilities for Regenerative Sequences. Informatics and Automation, Branching Processes and Related Topics, Tome 316 (2022), pp. 47-63. http://geodesic.mathdoc.fr/item/TRSPY_2022_316_a4/