Scalable Dynamic Mixture Model with Full Covariance for Probabilistic Traffic Forecasting

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Choi, Seongjin, Saunier, Nicolas, Zheng, Vincent Zhihao, Trepanier, Martin, Sun, Lijun
Natura: Preprint
Pubblicazione: 2022
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910001228939264
author Choi, Seongjin
Saunier, Nicolas
Zheng, Vincent Zhihao
Trepanier, Martin
Sun, Lijun
author_facet Choi, Seongjin
Saunier, Nicolas
Zheng, Vincent Zhihao
Trepanier, Martin
Sun, Lijun
contents Deep learning-based multivariate and multistep-ahead traffic forecasting models are typically trained with the mean squared error (MSE) or mean absolute error (MAE) as the loss function in a sequence-to-sequence setting, simply assuming that the errors follow an independent and isotropic Gaussian or Laplacian distributions. However, such assumptions are often unrealistic for real-world traffic forecasting tasks, where the probabilistic distribution of spatiotemporal forecasting is very complex with strong concurrent correlations across both sensors and forecasting horizons in a time-varying manner. In this paper, we model the time-varying distribution for the matrix-variate error process as a dynamic mixture of zero-mean Gaussian distributions. To achieve efficiency, flexibility, and scalability, we parameterize each mixture component using a matrix normal distribution and allow the mixture weight to change and be predictable over time. The proposed method can be seamlessly integrated into existing deep-learning frameworks with only a few additional parameters to be learned. We evaluate the performance of the proposed method on a traffic speed forecasting task and find that our method not only improves model performance but also provides interpretable spatiotemporal correlation structures.
format Preprint
id arxiv_https___arxiv_org_abs_2212_06653
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Scalable Dynamic Mixture Model with Full Covariance for Probabilistic Traffic Forecasting
Choi, Seongjin
Saunier, Nicolas
Zheng, Vincent Zhihao
Trepanier, Martin
Sun, Lijun
Machine Learning
Deep learning-based multivariate and multistep-ahead traffic forecasting models are typically trained with the mean squared error (MSE) or mean absolute error (MAE) as the loss function in a sequence-to-sequence setting, simply assuming that the errors follow an independent and isotropic Gaussian or Laplacian distributions. However, such assumptions are often unrealistic for real-world traffic forecasting tasks, where the probabilistic distribution of spatiotemporal forecasting is very complex with strong concurrent correlations across both sensors and forecasting horizons in a time-varying manner. In this paper, we model the time-varying distribution for the matrix-variate error process as a dynamic mixture of zero-mean Gaussian distributions. To achieve efficiency, flexibility, and scalability, we parameterize each mixture component using a matrix normal distribution and allow the mixture weight to change and be predictable over time. The proposed method can be seamlessly integrated into existing deep-learning frameworks with only a few additional parameters to be learned. We evaluate the performance of the proposed method on a traffic speed forecasting task and find that our method not only improves model performance but also provides interpretable spatiotemporal correlation structures.
title Scalable Dynamic Mixture Model with Full Covariance for Probabilistic Traffic Forecasting
topic Machine Learning
url https://arxiv.org/abs/2212.06653