Level crossings of the empirical process (Q1201897): Difference between revisions

Let \(\{U_ n(t);\;0\leq t\leq 1\}\) be the empirical process based on i.i.d. \(U(0,1)\) data. A generalization of the known asymptotics for the number of times the empirical distribution function crosses the true distribution function is obtained in terms of the crossing process \[ C^ x_ t(U_ n)=\#\{s\leq t;\;U_ n(s)=x\}. \] It is proven that the doubly-indexed stochastic process \(\{n^{-1/2}C^ x_ t(U_ n)\); \((x,t)\in R\times[0,1]\}\) converges in distribution (in the appropriate sense) to the Brownian bridge local time. A uniform strong approximation theorem for local times defined on one probability space is also proven.