On Simulations from the Two-Parameter Poisson-Dirichlet Process and the Normalized Inverse-Gaussian Process

TL;DR

This paper introduces efficient sampling procedures for the two-Parameter Poisson-Dirichlet and normalized inverse-Gaussian processes, demonstrating significant improvements over existing stick-breaking methods.

Contribution

The authors develop and compare new sampling approximations, showing enhanced efficiency over traditional stick-breaking approaches.

Findings

New sampling procedures are more efficient than existing methods.

Significant improvement in approximation accuracy.

Applicable to Bayesian nonparametric models.

Abstract

In this paper, we develop simple, yet efficient, procedures for sampling approximations of the two-Parameter Poisson-Dirichlet Process and the normalized inverse-Gaussian process. We compare the efficiency of the new approximations to the corresponding stick-breaking approximations, in which we demonstrate a substantial improvement.