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

DOI10.3150/22-BEJ1536arXiv1905.10733OpenAlexW2944950738MaRDI QIDQ6103233FDOQ6103233

Authors: Juho Lee, Xenia Miscouridou, Francois Caron 

Publication date: 2 June 2023

Published in: Bernoulli (Search for Journal in Brave)

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.

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)

Cites Work

• Modeling with normalized random measure mixture models
• Slice sampling. (With discussions and rejoinder)
• Inference from iterative simulation using multiple sequences
• Non-Gaussian Ornstein-Uhlenbeck-based models and some of their uses in financial economics. (With discussion)
• Title not available (Why is that?)
• Financial Modelling with Jump Processes
• A Representation of Independent Increment Processes without Gaussian Components
• Tempering stable processes
• Sampling the Dirichlet Mixture Model with Slices
• Completely random measures
• Particle Markov Chain Monte Carlo Methods
• Variational inference for Dirichlet process mixtures
• Hierarchical Mixture Modeling With Normalized Inverse-Gaussian Priors
• MCMC for normalized random measure mixture models
• Poisson/gamma random field models for spatial statistics
• Exact and approximate sum representations for the Dirichlet process
• Slice sampling mixture models
• Nonparametric Bayes estimators based on beta processes in models for life history data
• Size-biased sampling of Poisson point processes and excursions
• Distributional results for means of normalized random measures with independent increments
• Normalized random measures driven by increasing additive processes
• On simulation and properties of the stable law
• Posterior Analysis for Normalized Random Measures with Independent Increments
• Title not available (Why is that?)
• Gibbs Sampling Methods for Stick-Breaking Priors
• Generalized Gamma measures and shot-noise Cox processes
• Title not available (Why is that?)
• Title not available (Why is that?)
• On a class of Bayesian nonparametric estimates: I. Density estimates
• Survival models for heterogeneous populations derived from stable distributions
• Hidden Markov Models With Stick-Breaking Priors
• On series representations of infinitely divisible random vectors
• A blocked Gibbs sampler for NGG-mixture models via a priori truncation
• Controlling the Reinforcement in Bayesian Non-Parametric Mixture Models
• Bayesian Poisson process partition calculus with an application to Bayesian Lévy moving averages
• On the stick-breaking representation of \(\sigma\)-stable Poisson-Kingman models
• On simulation from infinitely divisible distributions
• Title not available (Why is that?)
• Title not available (Why is that?)
• Posterior sampling from \(\epsilon\)-approximation of normalized completely random measure mixtures
• Truncated random measures
• A moment-matching Ferguson \& Klass algorithm
• Sequential Monte Carlo methods for mixtures with normalized random measures with independent increments priors
• Sparse Graphs Using Exchangeable Random Measures
• Exchangeable trait allocations
• Bayesian Poisson calculus for latent feature modeling via generalized Indian buffet process priors
• Nonnegative Bayesian nonparametric factor models with completely random measures
• Importance conditional sampling for Pitman-Yor mixtures

Cited In (2)

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

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)