arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2605.21241cs.LG
收藏

划分与对比:无需增强学习鲁棒的时间特征

Divide and Contrast: Learning Robust Temporal Features without Augmentation

Department of Computer Science, Norwegian University of Science(挪威科学与技术大学计算机科学系);Department of Computer Science, United States Naval Academy(美国海军学院计算机科学系)

Abdul-Kazeem Shamba, Kerstin Bach, Gavin Taylor

AI总结:

本文提出Di-COT框架,通过对比时间窗口内的信息子结构而非单个时间步,实现了无需数据增强和多编码器传递的自监督学习,从而在六个大规模真实世界数据集和UCR/UEA基准上取得了最先进的性能,同时显著减少了训练时间。

摘要:

针对时间序列表示的自监督学习旨在减少对标记数据的依赖,同时保持强大的下游性能,但许多现有方法存在计算成本高或依赖不适用于多样化时间动态的假设。在本工作中,我们引入了Divide and Contrast (Di-COT),一种无需数据增强和多次编码器传递的无监督框架,通过对比时间窗口内的信息子结构而非单个时间步来实现。Di-COT在每次迭代中随机将每个窗口划分为少量重叠的子块,从而实现高效且有意义的对比,同时减轻时间转换期间的假阳性。为进一步提高可扩展性,我们采用了一种对比目标,其计算依赖于批量大小和子块数量,使损失计算独立于序列长度。在六个大规模真实世界数据集以及UCR和UEA基准上的广泛实验表明,Di-COT学习了语义结构化且可迁移的表示,实现了分类、聚类、kNN和跨数据集转移任务上的最先进的性能,同时大幅减少了训练时间。源代码可在https://github.com/sfi-norwai/Di-COT上公开获取。

英文摘要:

Self-supervised learning for time-series representation aims to reduce reliance on labeled data while maintaining strong downstream performance, yet many existing approaches incur high computational costs or rely on assumptions that do not hold across diverse temporal dynamics. In this work, we introduce Divide and Contrast (Di-COT), an unsupervised framework that avoids data augmentation and multiple encoder passes by contrasting informative substructures within a window rather than individual timesteps. Di-COT stochastically partitions each window into a small number of overlapping sub-blocks per iteration, enabling efficient and meaningful contrast while mitigating false positives during temporal transitions. To further improve scalability, we adopt a contrastive objective whose computation depends on the batch size and the number of sub-blocks, making loss computation independent of sequence length. Extensive experiments on six large-scale real-world datasets, as well as the UCR and UEA benchmarks, demonstrate that Di-COT learns semantically structured and transferable representations, achieving state-of-the-art performance on classification, clustering, $k$NN, and cross-dataset transfer, while substantially reducing training time. The source code is publicly available at https://github.com/sfi-norwai/Di-COT.

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。

↑