Add 2025 cooperative multi-agent anomaly detection publication

Author: dongwooky

content/publication/2025-cooperative-multi-agent-anomaly-detection-through-consensus-and-integration/index.md

@@ -0,0 +1,38 @@
+---
+title: Cooperative Multi-Agent Anomaly Detection through Consensus and Integration
+authors:
+- dongwook-kwon
+- hanwoong-ryu
+- jihwan-won
+- cheolsoo-park
+date: '2025-01-01T00:00:00Z'
+publishDate: '2025-01-01T00:00:00Z'
+doi: 10.5573/JSTS.2024.VV.X.1
+publication_types:
+- article-journal
+publication: IEIE Transactions on Smart Processing and Computing
+publication_short: IEIE TSPC
+abstract: >
+  Anomaly detection based on multivariate time series is essential for safety-critical
+  domains, where early identification of abnormal patterns ensures system reliability.
+  Traditional single-agent models and static ensembles often fail to capture complex
+  interchannel dependencies and dynamic anomaly propagation. To address these limitations,
+  this study proposes a multi-agent collaborative anomaly detection framework that
+  integrates Large Language Models and distributed intelligence. The framework consists
+  of hierarchical agent types including base detectors, consensus agents, ensemble
+  agents, and a master integrator coordinated through trust-based communication and
+  adaptive selection mechanisms. Experiments on the Mars Science Laboratory dataset
+  demonstrate substantial improvements over individual agents, particularly in recall
+  and F1 score, confirming the effectiveness of adaptive multi-agent cooperation.
+summary: ''
+url_pdf: https://doi.org/10.5573/JSTS.2024.VV.X.1
+tags:
+- Anomaly Detection
+- Multivariate Time Series
+- Multi-Agent Systems
+- Large Language Models
+featured: false
+draft: false
+projects: []
+slides: ''
+---