Precise asymptotics in the law of the logarithm for random fields in Hilbert space

From MaRDI portal
Publication:2644597





For \(d\geq 1\), let \(X, X_K;\;K\in {\mathbb Z}_{+}^d\) be i.i.d. random variables with \(EX=0\) taking values in a real separable Hilbert space, set \(S_N=\sum_{K\leq N} X_K,\;N\in {\mathbb Z}_{+}^d\), assume that \(E[\| X\|^2(\log\| X\|)^{b+d+3}]< \infty\) for some \(b>-d\), and put \(\sigma^2=\sup\{E[<X,y>^2]:\| y\|\leq 1\}\). For \(d\geq 2\), the authors compute the limit of \(\sum_N \frac{(\log| N| )^b}{| N|} P (\| S_{N}\|\geq \varepsilon\sigma\sqrt{| N| \log| N|}) \) as \(\varepsilon\to 0\). Then, for \(d=1\), they are interested in the almost sure behavior of \(\sum_N\frac{(\log| N| )^b}{| N|} I (\| S_{N}\|\geq\varepsilon\sigma\sqrt{| N| \log| N|})\) as \(\varepsilon\to 0\). Curiosly enough, strange statements of the form \(\lim_{\varepsilon\to 0}f_{\varepsilon}\leq c\) in \(L_2\) are encountered.











This page was built for publication: Precise asymptotics in the law of the logarithm for random fields in Hilbert space

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2644597)