Poisson--Gamma Dynamical Systems
Abstract
We introduce a new dynamical system for sequentially observed multivariate count data. This model is based on the gamma--Poisson construction---a natural choice for count data---and relies on a novel Bayesian nonparametric prior that ties and shrinks the model parameters, thus avoiding overfitting. We present an efficient MCMC inference algorithm that advances recent work on augmentation schemes for inference in negative binomial models. Finally, we demonstrate the model's inductive bias using a variety of real-world data sets, showing that it exhibits superior predictive performance over other models and infers highly interpretable latent structure.
Cite
@article{arxiv.1701.05573,
title = {Poisson--Gamma Dynamical Systems},
author = {Aaron Schein and Mingyuan Zhou and Hanna Wallach},
journal= {arXiv preprint arXiv:1701.05573},
year = {2017}
}
Comments
Appeared in the Proceedings of the 29th Advances in Neural Information Processing Systems (NIPS 2016)