Truncation-free Hybrid Inference for DPMM

2017/01/13 by Arnim Bleier, Bleier, Arnim
Computer Science · Mathematics · #Bayesian Methods and Mixture Models #FOS: Computer and information sciences #Gaussian Processes and Bayesian Inference #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Statistical Methods and Bayesian Inference

paper · pdf · doi:10.48550/arxiv.1701.03743

openalex publication_date 2017/01/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Dirichlet process mixture models (DPMM) are a cornerstone of Bayesian non-parametrics. While these models free from choosing the number of components a-priori, computationally attractive variational inference often reintroduces the need to do so, via a truncation on the variational distribution. In this paper we present a truncation-free hybrid inference for DPMM, combining the advantages of sampling-based MCMC and variational methods. The proposed hybridization enables more efficient variational updates, while increasing model complexity only if needed. We evaluate the properties of the hybrid updates and their empirical performance in single- as well as mixed-membership models. Our method is easy to implement and performs favorably compared to existing schemas.

Citations

• Practical Collapsed Stochastic Variational Inference for the HDP

Related

• Truncation Approximation for Enriched Dirichlet Process Mixture Models
• Simple approximate MAP Inference for Dirichlet processes
• Mixture models with a prior on the number of components
• Robust Model Fitting for Two-Part Model Within Bayesian Semiparametric Framework: Variational Approach
• Distributed Collapsed Gibbs Sampler for Dirichlet Process Mixture Models in Federated Learning