An extended urn model with application to approximation

From MaRDI portal





The author studies an extended version of Pólya's urn model, where a ball is drawn for each trial from an urn, initially containing some white and black balls, and a different number of balls of the color chosen and a different number of balls of the opposite color are added to the urn after each distinct trial. A class of general distributions and approximation operators with \(2n\) parameters is obtained. First, the author introduces the urn model and studies its basic distribution properties. Then, this urn model is applied to introduce and select a class of approximation operators for which the Weierstrass approximation theorem holds. Although the Weierstrass approximation theorem fails to hold for the distributions constructed by the Pólya-Eggenberger urn model for general functions, the model considered in the article allows more freedom to choose parameters. It makes possible to construct a sequence of some linear operators \({\mathcal L}_n\left[g\right]\) for a function \(g\in C\left(I\right)\), \(I\supset\left[0,1\right]\), based on the distribution considered and to prove its uniform convergence to the function \(g\) on \(\left[0,1\right]\). The article is recommended for the specialists in probability theory as well as in approximation theory.











This page was built for publication: An extended urn model with application to approximation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4813817)