Article ID Journal Published Year Pages File Type
1153573 Statistics & Probability Letters 2011 5 Pages PDF
Abstract
Random Bernstein polynomials induces a probability measure on the space of multivariate density functions on a unit cube. For density estimation, it is important that the Bernstein prior can be restricted to an admissible class of densities with certain geometric properties of the target density. In this article, we study the shape properties such as monotonicity, convexity, and symmetry of the Bernstein prior.
Related Topics
Physical Sciences and Engineering Mathematics Statistics and Probability
Authors
,