scientific article; zbMATH DE number 3901737

From MaRDI portal
Publication:3679995





Let \(U_ n\) be a U-statistic based on independent and nonidentically distributed random vectors. For i.i.d. random vectors, weak and strong invariance principles for \(U_ n\) have been obtained [cf. \textit{R. M. Loynes}, Proc. Am. Math. Soc. 25, 56-64 (1970; Zbl 0218.60046) and \textit{R. G. Miller} and the author, Ann. Math. Stat. 43, 31-41; (1972; Zbl 0238.62057)]. But, when \(X_ 1,...,X_ n\) are independent and nonidentically distributed random vectors, nothing has been done on the invariance principles since neither the basic reverse martingale property for \(U_ n\) nor the necessary martingale structure may hold. In this paper, the author established weak and strong invariance principles for \(U_ n\) based on independent and nonidentically distributed random vectors using a sequential decomposition for \(U_ n\) and some related martingale results.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3679995)