Abstract
A one-term Edgeworth expansion for U-statistics with kernel h(x, y) was derived by Jing and Wang [3] under optimal moment conditions. In this note, we show that one of the optimal moment conditions E| h(X 1, X 2|5/3 < ∞ can be weakened to lim t→∞ t 5/3 P(|h(X 1, X 2)| > t) → 0.
Similar content being viewed by others
REFERENCES
V. Bentkus and F. Gotze, On minimal moment assumptions in Berry-Esseen theorems for U-statistics, Theory Probab. Appl., 40, 430–445 (1995).
V. Bentkus, F. Gotze, and R. Zitikis, Lower estimates of the convergence rate for U-statistics, Ann. Probab., 22, 1707–1714 (1994).
B. Y. Jing and Q. Y. Wang, Edgeworth expansion for U-statistics under minimal conditions. Ann. Statist., 31, 1376–1391 (2003).
Author information
Authors and Affiliations
Additional information
Printed in Lietuvos Matematikos Rinkinys, Vol. 45, No. 3, pp. 453–440, July–September, 2005.
Rights and permissions
About this article
Cite this article
Jing, BY., Zhou, W. A Note on Edgeworth Expansions for U-Statistics under Minimal Conditions. Lith Math J 45, 353–358 (2005). https://doi.org/10.1007/s10986-005-0036-1
Received:
Issue Date:
DOI: https://doi.org/10.1007/s10986-005-0036-1