将多变量时序异常检测转为单变量均值变化检测，提升效率与可解释性。
TimePred: efficient and interpretable offline change point detection for high volume data -- with application to industrial process monitoring
- 通过预测时间索引将多变量问题简化为单变量均值变化检测
- 计算成本降低达两个数量级，性能媲美现有方法
- 支持特征级解释，适合工业过程监控场景
高维、大规模时序数据中的变化点检测在统计一致性、可扩展性和可解释性方面面临挑战。我们提出TimePred，一种自监督框架，通过预测每个样本的归一化时间索引，将多变量变化点检测转化为单变量均值变化检测。这使得可利用现有算法实现高效离线检测，并支持集成XAI归因方法进行特征级解释。实验表明，该方法在保持竞争力的同时，计算成本降低高达两个数量级。在工业制造案例研究中，检测准确率提升，验证了可解释变化点洞察的实际价值。
原文摘要 · Abstract (English)
Change-point detection (CPD) in high-dimensional, large-volume time series is challenging for statistical consistency, scalability, and interpretability. We introduce TimePred, a self-supervised framework that reduces multivariate CPD to univariate mean-shift detection by predicting each sample's normalized time index. This enables efficient offline CPD using existing algorithms and supports the integration of XAI attribution methods for feature-level explanations. Our experiments show competitive CPD performance while reducing computational cost by up to two orders of magnitude. In an industrial manufacturing case study, we demonstrate improved detection accuracy and illustrate the practical value of interpretable change-point insights.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。