Bayesian Multinomial Logistic Normal Models through Marginally Latent Matrix-T Processes

Justin D. Silverman, Kimberly Roche, Zachary C. Holmes, Lawrence A. David, Sayan Mukherjee
3/27/2019
stat.ME

Official code: jsilve24/stray (23 stars, last commit 2022-01-26).

Abstract

Bayesian multinomial logistic-normal (MLN) models are popular for the analysis of sequence count data (e.g., microbiome or gene expression data) due to their ability to model multivariate count data with complex covariance structure. However, existing implementations of MLN models are limited to handling small data sets due to the non-conjugacy of the multinomial and logistic-normal distributions. We introduce MLN models which can be written as marginally latent matrix-t process (LTP) models. Marginally LTP models describe a flexible class of generalized linear regression, non-linear regression, and time series models. We develop inference schemes for Marginally LTP models and, through application to MLN models, demonstrate that our inference schemes are both highly accurate and often 4-5 orders of magnitude faster than MCMC.

View on arXivView PDF

Code Implementations(1)

jsilve24/strayOfficial100%

Stray has been renamed fido, development continues at jsilve24/fido

235C, RJul 9, 20184 years ago

Similar Papers

Multinomial Logistic Regression: Asymptotic Normality on Null Covariates in High-Dimensions

81% match2023

Distributed multinomial regression

79% match1 repo92013

Bayesian Estimation of Mixed Multinomial Logit Models: Advances and Simulation-Based Evaluations

79% match1 repo72019

An Intuitive Tutorial to Gaussian Process Regression

78% match1 repo6952020

An Approximation Method for Fitted Random Forests

77% match2022

ParaRNN: An Interpretable and Parallelizable Recurrent Neural Network for Time-Dependent Data

77% match2026

Cite this paper

@article{silverman2019bayesian,
  title  = {Bayesian Multinomial Logistic Normal Models through Marginally Latent Matrix-T Processes},
  author = {Justin D. Silverman and Kimberly Roche and Zachary C. Holmes and Lawrence A. David and Sayan Mukherjee},
  year   = {2019},
  eprint = {1903.11695},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1903.11695}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1903.11695.svg)](https://codeofpaper.com/paper/1903.11695)

Discussion