Large and moderate deviations of empirical processes with nonstandard rates (Q1613084)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Large and moderate deviations of empirical processes with nonstandard rates |
scientific article |
Statements
Large and moderate deviations of empirical processes with nonstandard rates (English)
0 references
5 September 2002
0 references
The author discusses the large and moderate deviations of the type of empirical processes whose finite-dimensional distributions do not satisfy the Cramér condition. Nonstandard speeds and rate functions appear. He presents the following Theorem 1.1. Let \(\{\xi_{i,j}\}_{i,j=1}^\infty\) be a double sequence of symmetric i.i.d. r.v.'s with \({\mathbf P}(|\xi_{i,j}|\geq u)= e^{-u^p}\), for each \(u>0\), where \(0<p\leq 1\). Let \(\{b_n\}_{n=1}^\infty\) be a sequence of positive numbers such that \(n/b^{2-p}_n\to 0\). For each \(t\in T\), let \(\{x_j(t)\}_{j=1}^\infty\) be a sequence of real numbers such that \( \sum_{j=1}^{\infty}|x_j(t)|^2<\infty\). Then, the following conditions are equivalent: (a.1) \((T,d)\) is totally bounded, where \(d(s,t)=\sup_{j\geq 1} |x_j(s)-x_j(t)|\). (a.2) \(\sup_{t\in T}|X_n(t)|\to 0\operatorname {Pr}\), where \(X_n(t)= \frac{1}{b_n}\sum_{i=1}^{\infty} \sum_{j=1}^{\infty}x_j(t)\xi_{i,j}\). (b) \(\{X_n(t)\), \(t\in T\}\) satisfies the LDP with speed \(b^p_n\) and rate function \[ I(z)=\inf \Biggl\{\sum_{i=1}^{\infty}|\gamma_{i}|^p: \sum_{i=1}^{\infty} x_i(t)\gamma_i=z(t) \text{ for each } t\in T \Biggr\}. \]
0 references
large deviations
0 references
moderate deviations
0 references
empirical processes
0 references