Bayesian nonparametric modeling of latent partitions via Stirling-gamma priors

TL;DR

This paper introduces the Stirling-gamma prior for Dirichlet process precision, enhancing robustness and interpretability in Bayesian nonparametric models, with theoretical insights and ecological applications.

Contribution

The paper proposes a novel Stirling-gamma prior for Dirichlet process precision, providing analytical properties and conjugacy results that improve robustness and facilitate elicitation.

Findings

Stirling-gamma prior improves robustness of latent partition modeling.

Analytical results clarify properties of the induced partition.

Application demonstrates effectiveness in ecological community detection.

Abstract

Dirichlet process mixtures are particularly sensitive to the value of the precision parameter controlling the behavior of the latent partition. Randomization of the precision through a prior distribution is a common solution, which leads to more robust inferential procedures. However, existing prior choices do not allow for transparent elicitation, due to the lack of analytical results. We introduce and investigate a novel prior for the Dirichlet process precision, the Stirling-gamma distribution. We study the distributional properties of the induced random partition, with an emphasis on the number of clusters. Our theoretical investigation clarifies the reasons of the improved robustness properties of the proposed prior. Moreover, we show that, under specific choices of its hyperparameters, the Stirling-gamma distribution is conjugate to the random partition of a Dirichlet process. We illustrate with an ecological application the usefulness of our approach for the detection of communities of ant workers.