Enhancing Multivariate Time Series Forecasting via Multi-Task Learning and Random Matrix Theory
Abstract
We present a novel approach to multivariate time series forecasting by framing it as a multi-task learning problem. We propose an optimization strategy that enhances single-channel predictions by leveraging information across multiple channels. Our framework offers a closed-form solution for linear models and connects forecasting performance to key statistical properties using advanced analytical tools. Empirical results on both synthetic and real-world datasets demonstrate that integrating our method into training loss functions significantly improves univariate models by effectively utilizing multivariate data within a multi-task learning framework.
Cite
Text
Ilbert et al. "Enhancing Multivariate Time Series Forecasting via Multi-Task Learning and Random Matrix Theory." NeurIPS 2024 Workshops: TSALM, 2024.Markdown
[Ilbert et al. "Enhancing Multivariate Time Series Forecasting via Multi-Task Learning and Random Matrix Theory." NeurIPS 2024 Workshops: TSALM, 2024.](https://mlanthology.org/neuripsw/2024/ilbert2024neuripsw-enhancing/)BibTeX
@inproceedings{ilbert2024neuripsw-enhancing,
title = {{Enhancing Multivariate Time Series Forecasting via Multi-Task Learning and Random Matrix Theory}},
author = {Ilbert, Romain and Tiomoko, Malik and Louart, Cosme and Feofanov, Vasilii and Palpanas, Themis and Redko, Ievgen},
booktitle = {NeurIPS 2024 Workshops: TSALM},
year = {2024},
url = {https://mlanthology.org/neuripsw/2024/ilbert2024neuripsw-enhancing/}
}