Efficient Multivariate Time Series Forecasting via Calibrated Language Models with Privileged Knowledge Distillation

TL;DR

This paper introduces TimeKD, an efficient multivariate time series forecasting framework that leverages calibrated language models and privileged knowledge distillation to improve prediction quality while enhancing inference efficiency.

Contribution

The paper proposes a novel TimeKD framework combining calibrated language models and privileged knowledge distillation for efficient and accurate multivariate time series forecasting.

Findings

TimeKD achieves high-quality predictions with improved efficiency.

The cross-modality teacher model effectively captures future representations.

Extensive experiments demonstrate scalability and robustness of TimeKD.

Abstract

Multivariate time series forecasting (MTSF) endeavors to predict future observations given historical data, playing a crucial role in time series data management systems. With advancements in large language models (LLMs), recent studies employ textual prompt tuning to infuse the knowledge of LLMs into MTSF. However, the deployment of LLMs often suffers from low efficiency during the inference phase. To address this problem, we introduce TimeKD, an efficient MTSF framework that leverages the calibrated language models and privileged knowledge distillation. TimeKD aims to generate high-quality future representations from the proposed cross-modality teacher model and cultivate an effective student model. The cross-modality teacher model adopts calibrated language models (CLMs) with ground truth prompts, motivated by the paradigm of Learning Under Privileged Information (LUPI). In addition, we design a subtractive cross attention (SCA) mechanism to refine these representations. To cultivate an effective student model, we propose an innovative privileged knowledge distillation (PKD) mechanism including correlation and feature distillation. PKD enables the student to replicate the teacher's behavior while minimizing their output discrepancy. Extensive experiments on real data offer insight into the effectiveness, efficiency, and scalability of the proposed TimeKD.