Grasping Core Rules of Time Series through Pure Models

TL;DR

This paper introduces PureTS, a simple linear model that captures core time series rules, achieving state-of-the-art long-term forecasting with minimal complexity and high efficiency.

Contribution

Proposes PureTS, a pure linear neural network, demonstrating that simple models can effectively grasp time series core laws and outperform complex models in long-term forecasting.

Findings

PureTS achieves state-of-the-art results in 80% of long sequence tasks.

PureTS is nearly the lightest and fastest model for long-term prediction.

Pure linear layers help understand phenomena and improve prediction accuracy.

Abstract

Time series underwent the transition from statistics to deep learning, as did many other machine learning fields. Although it appears that the accuracy has been increasing as the model is updated in a number of publicly available datasets, it typically only increases the scale by several times in exchange for a slight difference in accuracy. Through this experiment, we point out a different line of thinking, time series, especially long-term forecasting, may differ from other fields. It is not necessary to use extensive and complex models to grasp all aspects of time series, but to use pure models to grasp the core rules of time series changes. With this simple but effective idea, we created PureTS, a network with three pure linear layers that achieved state-of-the-art in 80% of the long sequence prediction tasks while being nearly the lightest model and having the fastest running speed. On this basis, we discuss the potential of pure linear layers in both phenomena and essence. The ability to understand the core law contributes to the high precision of long-distance prediction, and reasonable fluctuation prevents it from distorting the curve in multi-step prediction like mainstream deep learning models, which is summarized as a pure linear neural network that avoids over-fluctuating. Finally, we suggest the fundamental design standards for lightweight long-step time series tasks: input and output should try to have the same dimension, and the structure avoids fragmentation and complex operations.