Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multivariate Time-series Anomaly Detection via Dynamic Model Pool & Ensembling

Wei Hu, Zewei Yu, Jianqiu Xu

TL;DR

This work tackles multivariate time-series anomaly detection by addressing limitations of single-model and fixed-dimension approaches. It introduces DMPEAD, a Dynamic Model Pool and Ensembling framework that builds a diverse, dimension-independent pool of basic models, updates it adaptively through a meta-model and merging, and ensembles a top-k subset for robust anomaly scoring. Key contributions include a parameter-transfer-based initial pool with a diversity-driven loss, an adaptive pool-update mechanism (expansion and merging), and a top-k, proxy-metric-based aggregation strategy for anomaly detection. Extensive experiments on 8 real-world datasets demonstrate strong performance, efficiency, and cross-dataset generalization, highlighting DMPEAD’s practical impact in scalable MTS anomaly detection.

Abstract

Multivariate time-series (MTS) anomaly detection is critical in domains such as service monitor, IoT, and network security. While multi-model methods based on selection or ensembling outperform single-model ones, they still face limitations: (i) selection methods rely on a single chosen model and are sensitive to the strategy; (ii) ensembling methods often combine all models or are restricted to univariate data; and (iii) most methods depend on fixed data dimensionality, limiting scalability. To address these, we propose DMPEAD, a Dynamic Model Pool and Ensembling framework for MTS Anomaly Detection. The framework first (i) constructs a diverse model pool via parameter transfer and diversity metric, then (ii) updates it with a meta-model and similarity-based strategy for adaptive pool expansion, subset selection, and pool merging, finally (iii) ensembles top-ranked models through proxy metric ranking and top-k aggregation in the selected subset, outputting the final anomaly detection result. Extensive experiments on 8 real-world datasets show that our model outperforms all baselines, demonstrating superior adaptability and scalability.

Multivariate Time-series Anomaly Detection via Dynamic Model Pool & Ensembling

TL;DR

This work tackles multivariate time-series anomaly detection by addressing limitations of single-model and fixed-dimension approaches. It introduces DMPEAD, a Dynamic Model Pool and Ensembling framework that builds a diverse, dimension-independent pool of basic models, updates it adaptively through a meta-model and merging, and ensembles a top-k subset for robust anomaly scoring. Key contributions include a parameter-transfer-based initial pool with a diversity-driven loss, an adaptive pool-update mechanism (expansion and merging), and a top-k, proxy-metric-based aggregation strategy for anomaly detection. Extensive experiments on 8 real-world datasets demonstrate strong performance, efficiency, and cross-dataset generalization, highlighting DMPEAD’s practical impact in scalable MTS anomaly detection.

Abstract

Multivariate time-series (MTS) anomaly detection is critical in domains such as service monitor, IoT, and network security. While multi-model methods based on selection or ensembling outperform single-model ones, they still face limitations: (i) selection methods rely on a single chosen model and are sensitive to the strategy; (ii) ensembling methods often combine all models or are restricted to univariate data; and (iii) most methods depend on fixed data dimensionality, limiting scalability. To address these, we propose DMPEAD, a Dynamic Model Pool and Ensembling framework for MTS Anomaly Detection. The framework first (i) constructs a diverse model pool via parameter transfer and diversity metric, then (ii) updates it with a meta-model and similarity-based strategy for adaptive pool expansion, subset selection, and pool merging, finally (iii) ensembles top-ranked models through proxy metric ranking and top-k aggregation in the selected subset, outputting the final anomaly detection result. Extensive experiments on 8 real-world datasets show that our model outperforms all baselines, demonstrating superior adaptability and scalability.
Paper Structure (13 sections, 4 equations, 6 figures, 4 tables)

This paper contains 13 sections, 4 equations, 6 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Problem Definition
  4. Our Method: DMPEAD
  5. Model Pool Training and Construction
  6. Dimension-independent Reconstruction Basic Model
  7. Initial Model Pool Construction
  8. Model Pool Testing and Updating
  9. Model Pool Expansion
  10. Model Pool Merging
  11. Model Ensembling and Anomaly Detection
  12. Experiments
  13. Conclusion

Figures (6)

  • Figure 1: An MTS example from server monitor, with anomalies in red
  • Figure 2: The proposed framework DMPEAD
  • Figure 3: Parameter sensitivity result on pool construction
  • Figure 4: Parameter sensitivity result on pool expansion
  • Figure 5: Parameter sensitivity result on pool merging
  • ...and 1 more figures