Parsimony or Capability? Decomposition Delivers Both in Long-term Time Series Forecasting

TL;DR

This paper shows that decomposition techniques enable long-term time series forecasting models to be both simpler and more effective, outperforming complex models with far fewer parameters by leveraging data's intrinsic dynamics.

Contribution

The study introduces a decomposition-based approach that reduces model complexity while maintaining or improving forecasting accuracy in long-term time series prediction.

Findings

Model outperforms benchmarks across datasets

Uses over 99% fewer parameters than competitors

Decomposition effectively controls model inflation

Abstract

Long-term time series forecasting (LTSF) represents a critical frontier in time series analysis, characterized by extensive input sequences, as opposed to the shorter spans typical of traditional approaches. While longer sequences inherently offer richer information for enhanced predictive precision, prevailing studies often respond by escalating model complexity. These intricate models can inflate into millions of parameters, resulting in prohibitive parameter scales. Our study demonstrates, through both analytical and empirical evidence, that decomposition is key to containing excessive model inflation while achieving uniformly superior and robust results across various datasets. Remarkably, by tailoring decomposition to the intrinsic dynamics of time series data, our proposed model outperforms existing benchmarks, using over 99 \% fewer parameters than the majority of competing methods. Through this work, we aim to unleash the power of a restricted set of parameters by capitalizing on domain characteristics--a timely reminder that in the realm of LTSF, bigger is not invariably better.