Poisson–Gamma Dynamical Systems

Conference paper

Aaron Schein, Mingyuan Zhou, Hanna M. Wallach
Advances in Neural Information Processing Systems (NeurIPS), 2016

View PDF

Cite

APA Click to copy
Schein, A., Zhou, M., & Wallach, H. M. (2016). Poisson–Gamma Dynamical Systems. In Advances in Neural Information Processing Systems (NeurIPS).

Chicago/Turabian Click to copy
Schein, Aaron, Mingyuan Zhou, and Hanna M. Wallach. “Poisson–Gamma Dynamical Systems.” In Advances in Neural Information Processing Systems (NeurIPS), 2016.

MLA Click to copy
Schein, Aaron, et al. “Poisson–Gamma Dynamical Systems.” Advances in Neural Information Processing Systems (NeurIPS), 2016.

BibTeX Click to copy

@inproceedings{aaron2016a,
  title = {Poisson–Gamma Dynamical Systems},
  year = {2016},
  author = {Schein, Aaron and Zhou, Mingyuan and Wallach, Hanna M.},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)}
}

Other materials: [Code] [Poster] (see below for video)

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.

This was selected for a full oral presentation (see below) among 8% of accepted papers!