计算机科学
异常检测
代表(政治)
多元统计
人工智能
背景(考古学)
水准点(测量)
模式识别(心理学)
时间序列
潜变量
编码(内存)
数据挖掘
系列(地层学)
机器学习
序列(生物学)
异常(物理)
上下文模型
潜变量模型
国家(计算机科学)
滑动窗口协议
可解释性
多元分析
作者
조동찬,Juwon Hwang,Keumyeong Kang,JiHo Han,Namsoon Jung
标识
DOI:10.48550/arxiv.2610.32337
摘要
In unsupervised multivariate time series anomaly detection (MTSAD), local variation can inflate reconstruction and forecasting errors, making observation-space discrepancy an unreliable indicator of abnormal system behavior. Since latent-space scoring can mitigate this limitation, we reframe one-class MTSAD as current-point system-state compatibility estimation and propose TENET, the Time-point Encoding Network for Multivariate Time Series Anomaly Detection. We define the current-point system state as a latent representation of the target time point, constructed from variable-level temporal context in the input window. TENET constructs this state through a temporal layer that retrieves window context relative to the current point, encodes time with current-point-centered embeddings, and adaptively mixes the retrieved context with the current representation. With this design, TENET treats the input window not as the object to reconstruct, forecast, summarize, or score, but as context for constructing a latent representation evaluated with a fixed standard-normal negative log-likelihood score. Across benchmark datasets, TENET achieves state-of-the-art performance. Additional experiments further show that TENET remains effective under alternative one-class objectives and outperforms substitutions based on generic sequence encoders, suggesting that its advantage comes from decision-aligned representation construction rather than from a specialized scoring rule. These results indicate that one-class MTSAD can be strong when the latent representation is designed around the current time point.
科研通智能强力驱动
Strongly Powered by AbleSci AI