IMTS is Worth Time $\times$ Channel Patches: Visual Masked Autoencoders for Irregular Multivariate Time Series Prediction

TL;DR

This paper introduces VIMTS, a novel framework that adapts visual Mask AutoEncoders for irregular multivariate time series forecasting, effectively handling missing data and capturing temporal patterns.

Contribution

VIMTS is the first to adapt visual MAE for IMTS forecasting, incorporating patch processing, cross-channel dependency learning, and self-supervised training for improved accuracy.

Findings

VIMTS outperforms existing methods in IMTS forecasting tasks.

VIMTS demonstrates strong few-shot learning capabilities.

VIMTS effectively handles extensive missing data in IMTS.

Abstract

Irregular Multivariate Time Series (IMTS) forecasting is challenging due to the unaligned nature of multi-channel signals and the prevalence of extensive missing data. Existing methods struggle to capture reliable temporal patterns from such data due to significant missing values. While pre-trained foundation models show potential for addressing these challenges, they are typically designed for Regularly Sampled Time Series (RTS). Motivated by the visual Mask AutoEncoder's (MAE) powerful capability for modeling sparse multi-channel information and its success in RTS forecasting, we propose VIMTS, a framework adapting Visual MAE for IMTS forecasting. To mitigate the effect of missing values, VIMTS first processes IMTS along the timeline into feature patches at equal intervals. These patches are then complemented using learned cross-channel dependencies. Then it leverages visual MAE's capability in handling sparse multichannel data for patch reconstruction, followed by a coarse-to-fine technique to generate precise predictions from focused contexts. In addition, we integrate self-supervised learning for improved IMTS modeling by adapting the visual MAE to IMTS data. Extensive experiments demonstrate VIMTS's superior performance and few-shot capability, advancing the application of visual foundation models in more general time series tasks. Our code is available at https://github.com/WHU-HZY/VIMTS.