Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection
Dongchan Cho, Jiho Han, Keumyeong Kang, Minsang Kim, Honggyu Ryu, Namsoon Jung
Abstract
Real-world multivariate time series anomalies are rare and often unlabeled. Additionally, prevailing methods rely on increasingly complex architectures tuned to benchmarks, detecting only fragments of anomalous segments and overstating performance. In this paper, we introduce OracleAD, a simple and interpretable unsupervised framework for multivariate time series anomaly detection. OracleAD encodes each variable's past sequence into a single causal embedding to jointly predict the present time point and reconstruct the input window, effectively modeling temporal dynamics. These embeddings then undergo a self-attention mechanism to project them into a shared latent space and capture spatial relationships. These relationships are not static, since they are modeled by a property that emerges from each variable's temporal dynamics. The projected embeddings are aligned to a Stable Latent Structure (SLS) representing normal-state relationships. Anomalies are identified using a dual scoring mechanism based on prediction error and deviation from the SLS, enabling fine-grained anomaly diagnosis at each time point and across individual variables. Since any noticeable SLS deviation originates from embeddings that violate the learned temporal causality of normal data, OracleAD directly pinpoints the root-cause variables at the embedding level. OracleAD achieves state-of-the-art results across multiple real-world datasets and evaluation protocols, while remaining interpretable through SLS.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
2 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Cluster-Aware Causal Mixer for Online Anomaly Detection in Multivariate Time Series cites Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
- Cluster-Aware Causal Mixer for Online Anomaly Detection in Multivariate Time Series cites Anomaly Transformer: Time Series Anomaly Detection with Association Discrepancy
- TranAD: Deep Transformer Networks for Anomaly Detection in Multivariate Time Series Data cites Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
- CAAD: Causality-Aware Multivariate Time Series Anomaly Detection via Multi-Scale Alignment and Structural Causal Consistency cites Anomaly Detection in Time Series: A Comprehensive Evaluation
- CAAD: Causality-Aware Multivariate Time Series Anomaly Detection via Multi-Scale Alignment and Structural Causal Consistency cites Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
- CAAD: Causality-Aware Multivariate Time Series Anomaly Detection via Multi-Scale Alignment and Structural Causal Consistency cites iTransformer: Inverted Transformers Are Effective for Time Series Forecasting
- CAAD: Causality-Aware Multivariate Time Series Anomaly Detection via Multi-Scale Alignment and Structural Causal Consistency cites Anomaly Transformer: Time Series Anomaly Detection with Association Discrepancy
- Anomaly Transformer: Time Series Anomaly Detection with Association Discrepancy cites Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
Cited by top-tier papers2
- Cluster-Aware Causal Mixer for Online Anomaly Detection in Multivariate Time SeriesMd Mahmuddun Nabi Murad, Yasin YilmazICML 2026 · 2 citations
- CAAD: Causality-Aware Multivariate Time Series Anomaly Detection via Multi-Scale Alignment and Structural Causal ConsistencyXin Wang, Yunshi Wen, Yanan He, Haotian Xu et al.KDD 2026
Builds on17
- iTransformer: Inverted Transformers Are Effective for Time Series ForecastingYong Liu, Tengge Hu, Haoran Zhang, Haixu Wu et al.ICLR 2024 · 1,703 citations
- Graph Neural Network-Based Anomaly Detection in Multivariate Time SeriesAilin Deng, Bryan HooiAAAI 2021 · 1,306 citations
- Anomaly Transformer: Time Series Anomaly Detection with Association DiscrepancyJiehui Xu, Haixu Wu, Jianmin Wang, Mingsheng LongICLR 2022 · 960 citations
- TranAD: Deep Transformer Networks for Anomaly Detection in Multivariate Time Series DataShreshth Tuli, Giuliano Casale, Nicholas R. JenningsVLDB 2022 · 930 citations
- Anomaly Detection in Time Series: A Comprehensive EvaluationSebastian Schmidl, Phillip Wenig, Thorsten PapenbrockVLDB 2022 · 578 citations
Related papers
- GCAD: Anomaly Detection in Multivariate Time Series from the Perspective of Granger CausalityZehao Liu, Mengzhou Gao, Pengfei JiaoAAAI 2025 · 13 citations
- Low Rank Transformer for Multivariate Time Series Anomaly Detection and LocalizationCharalampos Shimillas, Kleanthis Malialis, Konstantinos Fokianos, Marios M. PolycarpouICLR 2026
- Prototype-oriented unsupervised anomaly detection for multivariate time seriesYuxin Li, Wenchao Chen, Bo Chen, Dongsheng Wang et al.ICML 2023 · 31 citations
- T-Rep: Representation Learning for Time Series using Time-EmbeddingsArchibald Fraikin, Adrien Bennetot, Stéphanie AllassonnièreICLR 2024 · 24 citations
- SARAD: Spatial Association-Aware Anomaly Detection and Diagnosis for Multivariate Time SeriesZhihao Dai, Ligang He, Shuanghua Yang, Matthew LeekeNeurIPS 2024 · 27 citations
