Unlocking the Potential of Linear Networks for Irregular Multivariate Time Series Forecasting

TL;DR

This paper introduces AiT, an adaptive linear network with a Transformer module, designed to improve irregular multivariate time series forecasting by dynamically adjusting weights and capturing variable correlations, outperforming existing methods.

Contribution

The paper presents AiT, a novel adaptive linear network that handles irregular time series with dynamic weights and a Transformer module for variable correlation modeling, addressing key challenges in the field.

Findings

AiT improves prediction accuracy by 11% over state-of-the-art methods.

AiT reduces runtime by 52% compared to existing approaches.

Experiments on four benchmark datasets validate AiT's effectiveness.

Abstract

Time series forecasting holds significant importance across various industries, including finance, transportation, energy, healthcare, and climate. Despite the widespread use of linear networks due to their low computational cost and effectiveness in modeling temporal dependencies, most existing research has concentrated on regularly sampled and fully observed multivariate time series. However, in practice, we frequently encounter irregular multivariate time series characterized by variable sampling intervals and missing values. The inherent intra-series inconsistency and inter-series asynchrony in such data hinder effective modeling and forecasting with traditional linear networks relying on static weights. To tackle these challenges, this paper introduces a novel model named AiT. AiT utilizes an adaptive linear network capable of dynamically adjusting weights according to observation time points to address intra-series inconsistency, thereby enhancing the accuracy of temporal dependencies modeling. Furthermore, by incorporating the Transformer module on variable semantics embeddings, AiT efficiently captures variable correlations, avoiding the challenge of inter-series asynchrony. Comprehensive experiments across four benchmark datasets demonstrate the superiority of AiT, improving prediction accuracy by 11% and decreasing runtime by 52% compared to existing state-of-the-art methods.