Bayesian inference, model selection and likelihood estimation using fast rejection sampling: the Conway-Maxwell-Poisson distribution
From MaRDI portal
Publication:6201432
DOI10.1214/20-BA1230arXiv1709.03471MaRDI QIDQ6201432FDOQ6201432
Authors: A. P. Benson, Nial Friel 
Publication date: 20 February 2024
Published in: Bayesian Analysis (Search for Journal in Brave)
Abstract: Bayesian inference for models with intractable likelihood functions represents a challenging suite of problems in modern statistics. In this work we analyse the Conway-Maxwell-Poisson (COM-Poisson) distribution, a two parameter generalisation of the Poisson distribution. COM-Poisson regression modelling allows the flexibility to model dispersed count data as part of a generalised linear model (GLM) with a COM-Poisson response, where exogenous covariates control the mean and dispersion level of the response. The major difficulty with COM-Poisson regression is that the likelihood function contains multiple intractable normalising constants and is not amenable to standard inference and MCMC techniques. Recent work by Chanialidis et al. (2017) has seen the development of a sampler to draw random variates from the COM-Poisson likelihood using a rejection sampling algorithm. We provide a new rejection sampler for the COM-Poisson distribution which significantly reduces the CPU time required to perform inference for COM-Poisson regression models. A novel extension of this work shows that for any intractable likelihood function with an associated rejection sampler it is possible to construct unbiased estimators of the intractable likelihood which proves useful for model selection or for use within pseudo-marginal MCMC algorithms (Andrieu and Roberts, 2009). We demonstrate all of these methods on a real-world dataset of takeover bids.
Full work available at URL: https://arxiv.org/abs/1709.03471
Cites Work
- A flexible regression model for count data
- Estimating the dimension of a model
- Bayesian Measures of Model Complexity and Fit
- Negative Binomial Regression
- Multivariate output analysis for Markov chain Monte Carlo
- A Useful Distribution for Fitting Discrete Data: Revival of the Conway–Maxwell–Poisson Distribution
- The pseudo-marginal approach for efficient Monte Carlo computations
- Adaptive Rejection Sampling for Gibbs Sampling
- Exact and Computationally Efficient Likelihood-Based Estimation for Discretely Observed Diffusion Processes (with Discussion)
- Rao-Blackwellisation of sampling schemes
- An efficient Markov chain Monte Carlo method for distributions with intractable normalising constants
- Exact sampling with coupled Markov chains and applications to statistical mechanics
- Computer Generation of Poisson Deviates from Modified Normal Distributions
- On Russian roulette estimates for Bayesian inference with doubly-intractable likelihoods
- Noisy Monte Carlo: convergence of Markov chains with approximate transition kernels
- The BARISTA: a model for bid arrivals in online auctions
- Unbiased Bayesian inference for population Markov jump processes via random truncations
- Efficient Bayesian inference for COM-Poisson regression models
- Data augmentation for models based on rejection sampling
Cited In (4)
- Bayesian sparse vector autoregressive switching models with application to human gesture phase segmentation
- Generalized Bayesian Inference for Discrete Intractable Likelihood
- Characterizing Existence and Location of the ML Estimate in the Conway–Maxwell–Poisson Model
- Multivariate Conway-Maxwell-Poisson Distribution: Sarmanov Method and Doubly Intractable Bayesian Inference
This page was built for publication: Bayesian inference, model selection and likelihood estimation using fast rejection sampling: the Conway-Maxwell-Poisson distribution
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6201432)
