English

Better Batch for Deep Probabilistic Time Series Forecasting

Machine Learning 2024-10-22 v5 Machine Learning

Abstract

Deep probabilistic time series forecasting has gained attention for its ability to provide nonlinear approximation and valuable uncertainty quantification for decision-making. However, existing models often oversimplify the problem by assuming a time-independent error process and overlooking serial correlation. To overcome this limitation, we propose an innovative training method that incorporates error autocorrelation to enhance probabilistic forecasting accuracy. Our method constructs a mini-batch as a collection of DD consecutive time series segments for model training. It explicitly learns a time-varying covariance matrix over each mini-batch, encoding error correlation among adjacent time steps. The learned covariance matrix can be used to improve prediction accuracy and enhance uncertainty quantification. We evaluate our method on two different neural forecasting models and multiple public datasets. Experimental results confirm the effectiveness of the proposed approach in improving the performance of both models across a range of datasets, resulting in notable improvements in predictive accuracy.

Keywords

Cite

@article{arxiv.2305.17028,
  title  = {Better Batch for Deep Probabilistic Time Series Forecasting},
  author = {Vincent Zhihao Zheng and Seongjin Choi and Lijun Sun},
  journal= {arXiv preprint arXiv:2305.17028},
  year   = {2024}
}

Comments

The 27th International Conference on Artificial Intelligence and Statistics (AISTATS 2024); We corrected a misleading notation in the published version and added a link to the code

R2 v1 2026-06-28T10:47:41.759Z