Probabilistic Hierarchical Forecasting with Deep Poisson Mixtures

Abstract

Hierarchical forecasting problems arise when time series compose a group structure that naturally defines aggregation and disaggregation coherence constraints for the predictions. In this work, we explore a new forecast representation, the Poisson Mixture Mesh (PMM), that can produce probabilistic, coherent predictions; it is compatible with the neural forecasting innovations, and defines simple aggregation and disaggregation rules capable of accommodating hierarchical structures, unknown during its optimization. We perform an empirical evaluation to compare the PMM to other methods on Australian domestic tourism data.

PDF NeurIPSW OpenReview Semantic Scholar

Cite

Text

Olivares et al. "Probabilistic Hierarchical Forecasting with Deep Poisson Mixtures." NeurIPS 2021 Workshops: DGMs_Applications, 2021.

Markdown

[Olivares et al. "Probabilistic Hierarchical Forecasting with Deep Poisson Mixtures." NeurIPS 2021 Workshops: DGMs_Applications, 2021.](https://mlanthology.org/neuripsw/2021/olivares2021neuripsw-probabilistic/)

BibTeX

@inproceedings{olivares2021neuripsw-probabilistic,
  title     = {{Probabilistic Hierarchical Forecasting with Deep Poisson Mixtures}},
  author    = {Olivares, Kin Gutierrez and Meetei, Nganba and Ma, Ruijun and Reddy, Rohan and Cao, Mengfei},
  booktitle = {NeurIPS 2021 Workshops: DGMs_Applications},
  year      = {2021},
  url       = {https://mlanthology.org/neuripsw/2021/olivares2021neuripsw-probabilistic/}
}