Moderate deviations for randomly perturbed dynamical systems

From MaRDI portal





Small \(\varepsilon \) asymptotics of a Markov chain defined as a randomly perturbed discrete time dynamical system \[ X_n^{\varepsilon} = f(X_{n-1}^{\varepsilon}) +\varepsilon \eta _n^{\varepsilon} (\omega , X_{n-1}^{\varepsilon}) \] with the same initial condition \(X_0^{\varepsilon} =X_0\in R\) is studied, where \(f\) is a bounded continuous real function, \(\eta ^{\varepsilon}_n\) is measurable for each \(\varepsilon \) and \(n\) and forms an i.i.d. zero mean sequence of random variables for each \(x\). The paper deals with asymptotic analysis, as \(\varepsilon \to 0\), of the centered and normalized process \[ Y^{\varepsilon , \alpha}=\frac {1}{\varepsilon ^{1-\alpha}}[X^{\varepsilon}-X^0],\quad 1/2<\alpha <1. \] Under some conditions the processes \(Y^{\varepsilon , 1/2}\) and \(Y^{\varepsilon ,1}\) obey the central limit theorem and the large deviation principle, respectively. For \(1/2<\alpha <1\), asymptotic behaviour called the moderate deviation principle is shown and explicit rate function is found. An application to a model of density dependent branching process is given.



Cites work


Cited in
(24)








This page was built for publication: Moderate deviations for randomly perturbed dynamical systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1593615)