Limit theorems for von Mises statistics of a measure preserving transformation (Q466892): Difference between revisions

\(U\)- and \(V\)-statistics are multiple sums of functions of random variables. They provide important extensions of simple averages and have many applications in statistics. On the other hand, sample means of i.i.d.\ variables are particularly simple examples of averages of strictly stationary variables to which the ergodic theorem applies. The purpose of the paper under review is to provide a link between the \(U\) resp.\ \(V\) setting and the corresponding general ergodic framework also for multiple sums. Through an in-depth investigation, the authors show, e.g., how well-known concepts like Hoeffding's projection or martingale approximations reappear in a measure-preserving transformation setup. As main results, it is shown that consistency and distributional convergence of simple sums known from ergodic theory also hold for multiple sums.