Revisiting the Efficacy of Signal Decomposition in AI-based Time Series Prediction

TL;DR

This paper critically examines the role of signal decomposition in AI-based time series prediction, revealing that improper data processing can lead to misleadingly inflated results and emphasizing the need for causal data handling.

Contribution

It uncovers widespread dataset processing errors in existing methods and demonstrates that proper causal data handling reduces the perceived benefits of signal decomposition.

Findings

Improper dataset processing with future label leakage inflates results.

Causal data processing diminishes the effectiveness of signal decomposition.

Potential universal error in current time series modeling practices.

Abstract

Time series prediction is a fundamental problem in scientific exploration and artificial intelligence (AI) technologies have substantially bolstered its efficiency and accuracy. A well-established paradigm in AI-driven time series prediction is injecting physical knowledge into neural networks through signal decomposition methods, and sustaining progress in numerous scenarios has been reported. However, we uncover non-negligible evidence that challenges the effectiveness of signal decomposition in AI-based time series prediction. We confirm that improper dataset processing with subtle future label leakage is unfortunately widely adopted, possibly yielding abnormally superior but misleading results. By processing data in a strictly causal way without any future information, the effectiveness of additional decomposed signals diminishes. Our work probably identifies an ingrained and universal error in time series modeling, and the de facto progress in relevant areas is expected to be revisited and calibrated to prevent future scientific detours and minimize practical losses.