Large deviations for a class of chaos expansions

From MaRDI portal





The first part of the paper gives a general approximating continuous transformation principle for large deviations, and the second one contains applications of this principle to obtain large deviations for random variables having chaos expansions of exponential type. Let \(W = \{W_ t, t \in T\}\), \(T=[0,1]\), be a standard Wiener process in a probability space \((\Omega,{\mathcal F},P)\). For \(m \geq 1\), let \(I_ m(h_ m)\) be the \(m\)-th multiple Wiener-Itô integer of the symmetric kernel \(h_ m \in L^ 2(T^ m)\). The main result of applications is as follows: Let \(X\) be a random variable having a chaos expansion of exponential type. For each \(\varepsilon > 0\) define \(X^ \varepsilon = \sum^ \infty_{m = 0} \varepsilon^{m/2} I_ m (h_ m)\). Then the family of Wiener functionals \(\{X^ \varepsilon; \varepsilon > 0\}\) satisfies the large deviations principle on \(\mathbb{R}\).











This page was built for publication: Large deviations for a class of chaos expansions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1337947)