A symmetrization-desymmetrization procedure for uniformly good approximation of expectations involving arbitrary sums of generalized U-statistics
Klass, Michael J. ; Nowicki, Krzysztof
Ann. Probab., Tome 28 (2000) no. 1, p. 1884-1907 / Harvested from Project Euclid
Let $\Phi$ be a symmetric function, nondecreasing on $[0,\infty)$ and satisfying a $\Delta_2$ growth condition, $(X_1, Y_1), (X_2, Y_2),\ldots,( X_n,Y_n)$ be independent random vectors such that (for each $l\leqi\leq n)$ either $Y_i =X_i$ or $Y_i$ is independent of all the other variates, and the marginal distributions of ${X_i}$ and ${Y_j}$ are otherwise arbitrary. Let ${f_ {ij}(x, y)}_{1\leq i,j\leq n$ be any array of real valued measurable functions.We present a method of obtaining the order of magnitude of ¶ [image] ¶ The proof employs a double symmetrization,introducing independent copies ${\tilde{X}_i,\tilde{Y}_j}$ of ${X_i,Y_j}$, and moving from summands of the form $f _{ij}(X_i, Y_j)$ to what we call $f_{ij}^(s)(X_i,Y_j,\tilde{X}_i,\tilde{Y}_j)$. Substitution of fixed constants $\tilde{x}_i$ and $\tilde{y}_ j$ for $\tilde{X}_ i$ and $\tilde{Y}_ j$ results in $f_{ij}^(s)(X_i,Y_j,\tilde{x}_i,\tilde{y}_j)$, which equals $f_{ij}(X_i,Y_j)$ adjusted by a sum of quantities of first order separately in ${X_i}$ and ${Y_j}$. Introducing further explicit first-order adjustments, call them $g_{1ij}(X_i ,\tilde\mathbf{x},\tilde\mathbf{y})$ and $g_{2ij}(Y_j,\tilde\mathbf{x},\tilde\mathbf{y})$, it is proved that ¶ [image] ¶ where the latter is an explicitly computable quantity. For any $\tilde\mathbf{x}^0$ and $\tilde\mathbf{y}^0$ which come within a factor of two of minimizing $\Phi(\mathbf{f}^(s), \mathbf{X,Y}\tilde\mathbf{x,y})$ it is shown that ¶ [image]
Publié le : 2000-10-14
Classification:  Generalized $U$-statistics,  symmetrization –desymmetrization,  expectations of functions of second-order sums,  60E15,  60F25,  60G50
@article{1019160512,
     author = {Klass, Michael J. and Nowicki, Krzysztof},
     title = {A symmetrization-desymmetrization procedure for uniformly good
		 approximation of expectations involving arbitrary sums of generalized
		 U-statistics},
     journal = {Ann. Probab.},
     volume = {28},
     number = {1},
     year = {2000},
     pages = { 1884-1907},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1019160512}
}
Klass, Michael J.; Nowicki, Krzysztof. A symmetrization-desymmetrization procedure for uniformly good
		 approximation of expectations involving arbitrary sums of generalized
		 U-statistics. Ann. Probab., Tome 28 (2000) no. 1, pp.  1884-1907. http://gdmltest.u-ga.fr/item/1019160512/