A unified construction for series representations and finite approximations of completely random measures

DOI10.3150/22-BEJ1536MaRDI QIDQ6103233OpenAlexFDO

Authors Juho Lee, Xenia Miscouridou, Francois Caron

Publication date 2 June 2023

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1905.10733

zbMATH Keywords

slice samplingBayesian nonparametricsPoisson random measurescompletely random measuresgeneralised gamma process

Mathematics Subject Classification ID

Bayesian inference (62F15) Nonparametric estimation (62G05) Random measures (60G57)

Abstract: Infinite-activity completely random measures (CRMs) have become important building blocks of complex Bayesian nonparametric models. They have been successfully used in various applications such as clustering, density estimation, latent feature models, survival analysis or network science. Popular infinite-activity CRMs include the (generalized) gamma process and the (stable) beta process. However, except in some specific cases, exact simulation or scalable inference with these models is challenging and finite-dimensional approximations are often considered. In this work, we propose a general and unified framework to derive both series representations and finite-dimensional approximations of CRMs. Our framework can be seen as an extension of constructions based on size-biased sampling of Poisson point process [Perman1992]. It includes as special cases several known series representations as well as novel ones. In particular, we show that one can get novel series representations for the generalized gamma process and the stable beta process. We also provide some analysis of the truncation error.

Recommendations

• Truncated random measures
• On the Truncation Error of a Superposed Gamma Process
• A moment-matching Ferguson \& Klass algorithm
• Posteriors, conjugacy, and exponential families for completely random measures
• Posterior sampling from \(\epsilon\)-approximation of normalized completely random measure mixtures

Cites work

• scientific article; zbMATH DE number 1639863 (Why is no real title available?)
• scientific article; zbMATH DE number 3518091 (Why is no real title available?)
• scientific article; zbMATH DE number 1219014 (Why is no real title available?)
• scientific article; zbMATH DE number 227027 (Why is no real title available?)
• scientific article; zbMATH DE number 5166032 (Why is no real title available?)
• scientific article; zbMATH DE number 3026523 (Why is no real title available?)
• A Representation of Independent Increment Processes without Gaussian Components
• A blocked Gibbs sampler for NGG-mixture models via a priori truncation
• A moment-matching Ferguson \& Klass algorithm
• Bayesian Poisson calculus for latent feature modeling via generalized Indian buffet process priors
• Bayesian Poisson process partition calculus with an application to Bayesian Lévy moving averages
• Completely random measures
• Controlling the Reinforcement in Bayesian Non-Parametric Mixture Models
• Distributional results for means of normalized random measures with independent increments
• Exact and approximate sum representations for the Dirichlet process
• Exchangeable trait allocations
• Financial Modelling with Jump Processes
• Generalized Gamma measures and shot-noise Cox processes
• Gibbs Sampling Methods for Stick-Breaking Priors
• Hidden Markov Models With Stick-Breaking Priors
• Hierarchical Mixture Modeling With Normalized Inverse-Gaussian Priors
• Importance conditional sampling for Pitman-Yor mixtures
• Inference from iterative simulation using multiple sequences
• MCMC for normalized random measure mixture models
• Modeling with normalized random measure mixture models
• Non-Gaussian Ornstein-Uhlenbeck-based models and some of their uses in financial economics. (With discussion)
• Nonnegative Bayesian nonparametric factor models with completely random measures
• Nonparametric Bayes estimators based on beta processes in models for life history data
• Normalized random measures driven by increasing additive processes
• On a class of Bayesian nonparametric estimates: I. Density estimates
• On series representations of infinitely divisible random vectors
• On simulation and properties of the stable law
• On simulation from infinitely divisible distributions
• On the stick-breaking representation of \(\sigma\)-stable Poisson-Kingman models
• Particle Markov Chain Monte Carlo Methods
• Poisson/gamma random field models for spatial statistics
• Posterior Analysis for Normalized Random Measures with Independent Increments
• Posterior sampling from \(\epsilon\)-approximation of normalized completely random measure mixtures
• Sampling the Dirichlet Mixture Model with Slices
• Sequential Monte Carlo methods for mixtures with normalized random measures with independent increments priors
• Size-biased sampling of Poisson point processes and excursions
• Slice sampling mixture models
• Slice sampling. (With discussions and rejoinder)
• Sparse graphs using exchangeable random measures
• Survival models for heterogeneous populations derived from stable distributions
• Tempering stable processes
• Truncated random measures
• Variational inference for Dirichlet process mixtures

Cited in

(2)

• On the uniform convergence of random series in Skorohod space and representations of càdlàg infinitely divisible processes
• On series representations of infinitely divisible random vectors

This page was built for publication: A unified construction for series representations and finite approximations of completely random measures

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6103233)