TY - JOUR
T1 - METER
T2 - 50th International Conference on Very Large Data Bases, VLDB 2024
AU - Zhu, Jiaqi
AU - Deng, Fang
AU - Cai, Shaofeng
AU - Ooi, Beng Chin
AU - Zhang, Wenqiao
N1 - Publisher Copyright:
© 2023, VLDB Endowment. All rights reserved.
PY - 2023
Y1 - 2023
N2 - Real-time analytics and decision-making require online anomaly detection (OAD) to handle drifts in data streams efficiently and effectively. Unfortunately, existing approaches are often constrained by their limited detection capacity and slow adaptation to evolving data streams, inhibiting their efficacy and efficiency in handling concept drift, which is a major challenge in evolving data streams. In this paper, we introduce METER, a novel dynamic concept adaptation framework that introduces a new paradigm for OAD. METER addresses concept drift by first training a base detection model on historical data to capture recurring central concepts, and then learning to dynamically adapt to new concepts in data streams upon detecting concept drift. Particularly, METER employs a novel dynamic concept adaptation technique that leverages a hypernetwork to dynamically generate the parameter shift of the base detection model, providing a more effective and efficient solution than conventional retraining or fine-tuning approaches. Further, METER incorporates a lightweight drift detection controller, underpinned by evidential deep learning, to support robust and interpretable concept drift detection. We conduct an extensive experimental evaluation, and the results show that METER significantly outperforms existing OAD approaches in various application scenarios.
AB - Real-time analytics and decision-making require online anomaly detection (OAD) to handle drifts in data streams efficiently and effectively. Unfortunately, existing approaches are often constrained by their limited detection capacity and slow adaptation to evolving data streams, inhibiting their efficacy and efficiency in handling concept drift, which is a major challenge in evolving data streams. In this paper, we introduce METER, a novel dynamic concept adaptation framework that introduces a new paradigm for OAD. METER addresses concept drift by first training a base detection model on historical data to capture recurring central concepts, and then learning to dynamically adapt to new concepts in data streams upon detecting concept drift. Particularly, METER employs a novel dynamic concept adaptation technique that leverages a hypernetwork to dynamically generate the parameter shift of the base detection model, providing a more effective and efficient solution than conventional retraining or fine-tuning approaches. Further, METER incorporates a lightweight drift detection controller, underpinned by evidential deep learning, to support robust and interpretable concept drift detection. We conduct an extensive experimental evaluation, and the results show that METER significantly outperforms existing OAD approaches in various application scenarios.
UR - http://www.scopus.com/inward/record.url?scp=85190649349&partnerID=8YFLogxK
U2 - 10.14778/3636218.3636233
DO - 10.14778/3636218.3636233
M3 - Conference article
AN - SCOPUS:85190649349
SN - 2150-8097
VL - 17
SP - 694
EP - 807
JO - Proceedings of the VLDB Endowment
JF - Proceedings of the VLDB Endowment
IS - 4
Y2 - 24 August 2024 through 29 August 2024
ER -