Some observations on the KMT dyadic scheme (Q866635)

Mason investigates the well-known Komlós-Major-Tusnády (KMT) Brownian bridge approximation to the uniform empirical process. He describes the KMT dyadic scheme and gives some remarks relating to the approaches of other authors working on this field. The author points out that the KMT construction is a Gaussian approximation to a nested array of dyadic multinomial random vectors. Using the Haar basis method he derives from the KMT approach an approximation result for the uniform empirical process, which holds uniformly over the class of dyadic cubes in \([0,1]^d\).