arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2203.09978cs.LGstat.ML
收藏

WOODS:时间序列中的分布外泛化基准

WOODS: Benchmarks for Out-of-Distribution Generalization in Time Series

Mila - Québec AI Institute(米拉-魁北克人工智能研究所);Gina Cody School of Engineering and Computer Science(吉娜·科迪工程与计算机科学学院);Concordia University(康考迪亚大学);CHU Sainte-Justine Research Center(圣贾斯汀大学医院研究中心)

Jean-Christophe Gagnon-Audet, Kartik Ahuja, Mohammad-Javad Darvishi-Bayazi, Pooneh Mousavi, Guillaume Dumas, Irina Rish

AI总结:

针对时间序列任务中分布外泛化研究不足的问题,提出包含8个开源基准的WOODS框架,评估并修订现有算法,揭示时间序列带来的新挑战。

摘要:

机器学习模型在分布偏移下通常难以良好泛化。理解和克服这些失败催生了分布外(OOD)泛化研究领域。尽管在静态计算机视觉任务中被广泛研究,但OOD泛化在时间序列任务中仍探讨不足。为揭示这一空白,我们提出WOODS:八个具有挑战性的开源时间序列基准,涵盖多种数据模态,如视频、脑记录和传感器信号。我们针对时间序列任务修订了现有的OOD泛化算法,并使用我们的系统框架对其进行评估。实验表明,经验风险最小化和OOD泛化算法在我们的数据集上仍有很大改进空间,从而凸显了时间序列任务带来的新挑战。代码和文档可在 https://woods-benchmarks.github.io 获取。

英文摘要:

Machine learning models often fail to generalize well under distributional shifts. Understanding and overcoming these failures have led to a research field of Out-of-Distribution (OOD) generalization. Despite being extensively studied for static computer vision tasks, OOD generalization has been underexplored for time series tasks. To shine light on this gap, we present WOODS: eight challenging open-source time series benchmarks covering a diverse range of data modalities, such as videos, brain recordings, and sensor signals. We revise the existing OOD generalization algorithms for time series tasks and evaluate them using our systematic framework. Our experiments show a large room for improvement for empirical risk minimization and OOD generalization algorithms on our datasets, thus underscoring the new challenges posed by time series tasks. Code and documentation are available at https://woods-benchmarks.github.io .

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。

↑