Mixture modeling via vectors of normalized independent finite point processes

Statistical modeling in presence of hierarchical data is a crucial task in Bayesian statistics. The Hierarchical Dirichlet Process (HDP) represents the utmost tool to handle data organized in groups through mixture modeling. Although the HDP is mathematically tractable, its computational cost is typically demanding, and its analytical complexity represents a barrier for practitioners. The present paper conceives a mixture model based on a novel family of Bayesian priors designed for multilevel data and obtained by normalizing a finite point process. A full distribution theory for this new family and the induced clustering is developed, including tractable expressions for marginal, posterior and predictive distributions. Efficient marginal and conditional Gibbs samplers are designed for providing posterior inference. The proposed mixture model overcomes the HDP in terms of analytical feasibility, clustering discovery, and computational time. The motivating application comes from the analysis of shot put data, which contains performance measurements of athletes across different seasons. In this setting, the proposed model is exploited to induce clustering of the observations across seasons and athletes. By linking clusters across seasons, similarities and differences in athlete's performances are identified.