Latent feature regression for multivariate count data
Arto Klami, Abhishek Tripathi, Johannes Sirola, Lauri Väre, Frederic Roulland
Abstract
We consider the problem of regression on multivariate count data and present a Gibbs sampler for a latent feature regression model suitable for both under- and overdispersed response variables. The model learns count-valued latent features conditional on arbitrary covariates, modeling them as negative binomial variables, and maps them into the dependent count-valued observations using a Dirichlet-multinomial distribution. From another viewpoint, the model can be seen as a generalization of a specific topic model for scenarios where we are interested in generating the actual counts of observations and not just their relative frequencies and co-occurrences. The model is demonstrated on a smart traffic application where the task is to predict public transportation volume for unknown locations based on a characterization of the close-by services and venues.
BibTeX
@InProceedings{pmlr-v38-klami15,
title = {{Latent feature regression for multivariate count data}},
author = {Klami, Arto and Tripathi, Abhishek and Sirola, Johannes and Väre, Lauri and Roulland, Frederic},
booktitle = {Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics},
pages = {462--470},
year = {2015},
editor = {Lebanon, Guy and Vishwanathan, S. V. N.},
volume = {38},
series = {Proceedings of Machine Learning Research},
address = {San Diego, California, USA},
month = {09--12 May},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v38/klami15.pdf},
url = {https://proceedings.mlr.press/v38/klami15.html},
abstract = {We consider the problem of regression on multivariate count data and present a Gibbs sampler for a latent feature regression model suitable for both under- and overdispersed response variables. The model learns count-valued latent features conditional on arbitrary covariates, modeling them as negative binomial variables, and maps them into the dependent count-valued observations using a Dirichlet-multinomial distribution. From another viewpoint, the model can be seen as a generalization of a specific topic model for scenarios where we are interested in generating the actual counts of observations and not just their relative frequencies and co-occurrences. The model is demonstrated on a smart traffic application where the task is to predict public transportation volume for unknown locations based on a characterization of the close-by services and venues.}
}