Small deviations of stable processes and entropy of the associated random operators

Abstract

We investigate the relation between the small deviation problem for a symmetric α-stable random vector in a Banach space and the metric entropy properties of the operator generating it. This generalizes former results due to Li and Linde and to Aurzada. It is shown that this problem is related to the study of the entropy numbers of a certain random operator. In some cases, an interesting gap appears between the entropy of the original operator and that of the random operator generated by it. This phenomenon is studied thoroughly for diagonal operators. Basic ingredients here are techniques related to random partitions of the integers. The main result concerning metric entropy and small deviations allows us to determine or provide new estimates for the small deviation rate for several symmetric α-stable random processes, including unbounded Riemann-Liouville processes, weighted Riemann-Liouville processes and the (d-dimensional) α-stable sheet. This is an electronic reprint of the original article published by the ISI/BS in Bernoulli, 2009, Vol. 15, No. 4, 1305-1334. This reprint differs from the original in pagination and typographic detail. 1350-7265 c 2009 ISI/BS 1306 F. Aurzada, M. Lifshits and W. LindeIn this case, we say that X is generated by the operator u. This approach is very useful for investigating symmetric α-stable random processes with paths in E ′ . We refer to Section 5 of Li and Linde (2004) or Section 7.1.1 below for a discussion of how all natural examples of SαS processes fit into this framework. For example, if u from L p [0, 1] to L α [0, 1] is defined by (uf )(t) :=

Small ball probabilities, metric entropy and Gaussian rough paths

“…The link between Small Ball Probabilities for Gaussian measures on Banach spaces is explored in [KL93a] and later extended in [LL99] to encompass the truncation of Gaussian measures. The same ideas to link the small ball probabilities and metric entropies for Stable processes are studied in [LL04,Aur07,ALL09]. Small Ball Probability results for Integrated Brownian motion, see [GHT03], were used to compute the Metric Entropy of k-monotone functions in [Gao08].…”

A general approach to small deviation via concentration of measures

“…Developing a general strategy to deal with the small deviation problem for Gaussian processes is culminated with giving a precise link, discovered by Kuelbs and Li [12] and completed by Li and Linde [16], to the metric entropy of the unit ball of the reproducing kernel Hilbert space generated by Gaussian process. In the non-Gaussian case, similar links are built in [17,1,3] for symmetric α-stable processes. Apparently, it remains a great challenge to find some principle describing small deviations for general classes of processes and norms, rather than investigate the problem case by case.…”

Small ball probabilities and large deviations for grey Brownian motion

“…For Gaussian processes, such as fBm (β = 1), the small ball problem has been studied extensively [13], and exponential decay is typical. But there are also many works studying small ball probabilities for non-Gaussian processes; see, e.g., [1,2] and the references therein. We refer to [11,18] for other examples of processes with the small ball rate ε 2 of ggBm.…”