Delayformer: spatiotemporal transformation for predicting high-dimensional dynamics

TL;DR

Delayformer introduces a novel spatiotemporal transformer framework that predicts high-dimensional system dynamics by transforming variables into delay-embedded states, effectively handling nonlinearity and complex interactions.

Contribution

It develops a multivariate spatiotemporal information transformation and a transformer-based architecture to improve high-dimensional time-series prediction under limited and noisy data.

Findings

Outperforms state-of-the-art methods on synthetic and real datasets.

Successfully predicts system states rather than individual variables.

Demonstrates broad applicability across different domains.

Abstract

Predicting time-series is of great importance in various scientific and engineering fields. However, in the context of limited and noisy data, accurately predicting dynamics of all variables in a high-dimensional system is a challenging task due to their nonlinearity and also complex interactions. Current methods including deep learning approaches often perform poorly for real-world systems under such circumstances. This study introduces the Delayformer framework for simultaneously predicting dynamics of all variables, by developing a novel multivariate spatiotemporal information (mvSTI) transformation that makes each observed variable into a delay-embedded state (vector) and further cross-learns those states from different variables. From dynamical systems viewpoint, Delayformer predicts system states rather than individual variables, thus theoretically and computationally overcoming such nonlinearity and cross-interaction problems. Specifically, it first utilizes a single shared Visual Transformer (ViT) encoder to cross-represent dynamical states from observed variables in a delay embedded form and then employs distinct linear decoders for predicting next states, i.e. equivalently predicting all original variables parallelly. By leveraging the theoretical foundations of delay embedding theory and the representational capabilities of Transformers, Delayformer outperforms current state-of-the-art methods in forecasting tasks on both synthetic and real-world datasets. Furthermore, the potential of Delayformer as a foundational time-series model is demonstrated through cross-domain forecasting tasks, highlighting its broad applicability across various scenarios.