Lognormal and Gamma Mixed Negative Binomial Regression

Abstract

In regression analysis of counts, a lack of simple and efficient algorithms for posterior computation has made Bayesian approaches appear unattractive and thus underdeveloped. We propose a lognormal and gamma mixed negative binomial (NB) regression model for counts, and present efficient closed-form Bayesian inference; unlike conventional Poisson models, the proposed approach has two free parameters to include two different kinds of random effects, and allows the incorporation of prior information, such as sparsity in the regression coefficients. By placing a gamma distribution prior on the NB dispersion parameter r, and connecting a lognormal distribution prior with the logit of the NB probability parameter p, efficient Gibbs sampling and variational Bayes inference are both developed. The closed-form updates are obtained by exploiting conditional conjugacy via both a compound Poisson representation and a Polya-Gamma distribution based data augmentation approach. The proposed Bayesian inference can be implemented routinely, while being easily generalizable to more complex settings involving multivariate dependence structures. The algorithms are illustrated using real examples.

Appears in Proceedings of the 29th International Conference on Machine Learning (ICML 2012)

References in corpus (3)

• Variational Relevance Vector Machines
• Elliptical slice sampling
• Default Bayesian analysis for multi-way tables: a data-augmentation approach

Cited by in corpus (17)

• Dependent Multinomial Models Made Easy: Stick Breaking with the Pólya-Gamma Augmentation
• A New Spatial Count Data Model with Bayesian Additive Regression Trees for Accident Hot Spot Identification
• Semi-Implicit Variational Inference
• Augment-and-Conquer Negative Binomial Processes
• Expectation-maximization for logistic regression
• PG-TS: Improved Thompson Sampling for Logistic Contextual Bandits
• Bayesian inference for logistic models using Polya-Gamma latent variables
• Softplus Regressions and Convex Polytopes
• SimCD: Simultaneous Clustering and Differential expression analysis for single-cell transcriptomic data
• Negative Binomial Matrix Factorization for Recommender Systems
• Fragmentation Coagulation Based Mixed Membership Stochastic Blockmodel
• The Hawkes Edge Partition Model for Continuous-time Event-based Temporal Networks
• Doubly Non-Central Beta Matrix Factorization for DNA Methylation Data
• Bayesian Gamma-Negative Binomial Modeling of Single-Cell RNA Sequencing Data
• Bayesian method for inferring the impact of geographical distance on intensity of communication
• MCMC-Interactive Variational Inference
• Coresets for Dependency Networks