MTS-DMAE: Dual-Masked Autoencoder for Unsupervised Multivariate Time Series Representation Learning

Abstract--Unsupervised multivariate time series (MTS) representation learning aims to extract compact and informative representations from raw sequences without relying on labels, enabling efficient transfer to diverse downstream tasks. In this paper, we propose Dual-Masked Autoencoder (DMAE), a novel masked time-series modeling framework for unsupervised MTS representation learning. DMAE formulates two complementary pretext tasks: (1) reconstructing masked values based on visible attributes, and (2) estimating latent representations of masked features, guided by a teacher encoder . T o further improve representation quality, we introduce a feature-level alignment constraint that encourages the predicted latent representations to align with the teacher's outputs. By jointly optimizing these objectives, DMAE learns temporally coherent and semantically rich representations. Comprehensive evaluations across classification, regression, and forecasting tasks demonstrate that our approach achieves consistent and superior performance over competitive baselines. Multivariate time series (MTS) data constitute a fundamental modality that appears in numerous application areas. Typically, MTS data describe the temporal evolution of multiple synchronized variables, such as simultaneous measurements of diverse physical quantities.