Inferential Statisticshard
0:00.0

In the context of the Bahadur-Rao theorem for large deviations, if Sn=i=1nXiS_n = \sum_{i=1}^n X_i with E[Xi]=μ<aE[X_i] = \mu < a, what is the asymptotic behavior of the tail probability P(Snna)P(S_n \geq na)?