文摘
Let \(N\) be a Poisson distributed random variable (r.v.) with parameter \(\lambda \). Let \(\{X, X_i, i \ge 1\}\) be a sequence of i.i.d. r.v’s that are independent of \(N\). Set \(S_N=\sum _{j=1}^NX_j\) and \(V_N^2=\sum _{j=1}^NX_j^2\). Assume that \(0<\mu =EX<\infty \) and \(EX^4 < \infty \). In this paper, it is proved that \(P(S_N-\lambda \mu \ge x V_N ) / \{1- \Phi (x)\} \rightarrow 1\) uniformly in \(x \in [0, o(\lambda ^{1/6}))\), as \(\lambda \rightarrow \infty \). Keywords Cramér large deviation Self-normalized compound Poisson sum Studentized compound Poisson sum