arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.00889cs.LG
收藏

评估噪声周期时间序列上的时间序列基础模型

Evaluating Time Series Foundation Models on Noisy Periodic Time Series

Syamantak Datta Gupta

AI总结:

本文通过两个噪声周期合成数据集,在不同噪声、频率和采样率下实证评估主流TSFM的零样本长视界预测,并与FFT和AR基准比较,揭示其性能边界。

摘要:

尽管基础模型的最新进展已显著影响机器学习,但对时间序列基础模型(TSFM)性能的严格测试在很大程度上仍未得到充分探索。本文提出一项实证研究,在两个由噪声周期时间序列构成的合成数据集上,评估若干领先TSFM的零样本、长视界预测能力。我们在不同噪声水平、基础频率和采样率下评估模型有效性。作为比较基准,我们选择两种统计技术:基于Fourier变换(FFT)的方法和线性自回归(AR)模型。研究结果表明,对于具有有界周期和较高采样率的时间序列,TSFM能够匹敌甚至优于统计方法;但随着周期更长、噪声水平更高、采样率更低以及时间序列形状更复杂,其预测能力会下降。

英文摘要:

While recent advancements in foundation models have significantly impacted machine learning, rigorous tests on the performance of time series foundation models (TSFMs) remain largely underexplored. This paper presents an empirical study evaluating the zero-shot, long-horizon forecasting abilities of several leading TSFMs over two synthetic datasets constituting noisy periodic time series. We assess model efficacy across different noise levels, underlying frequencies, and sampling rates. As benchmarks for comparison, we choose two statistical techniques: a Fourier transform (FFT)-based approach and a linear autoregressive (AR) model. Our findings demonstrate that while for time series with bounded periods and higher sampling rates, TSFMs can match or outperform the statistical approaches, their forecasting abilities deteriorate with longer periods, higher noise levels, lower sampling rates and more complex shapes of the time series.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑