Earticle

초록

영어

This paper addresses the critical task of anomaly detection in river network sensor data, essential for accurate and continuous water quality monitoring. We propose M-MAD (Multi-Modal Anomaly Detection), a novel approach that integrates multi-modal features, including sensor data, weather information, and historical anomalies. M-MAD builds on the Graph Deviation Network (GDN) framework by introducing an improved anomaly threshold criterion derived from the learned graph structure. Our evaluation employs rigorous benchmarking simulations that mimic complex dependency structures and diverse anomalies, thoroughly assessing the strengths and weaknesses of M-MAD compared to existing methods. Results demonstrate M-MAD's superior performance in handling high-dimensional datasets and its enhanced interpretability, crucial for effective anomaly detection.

목차

Abstract
I. INTRODUCTION
II. METHODOLOGY
A. Sensor Data Generation
B. Simulated Response Generation
C. Feature Integration
D. Anomaly Generation
III. EXPERIMENTAL RESULTS
A. Data Splitting
B. Forecasting-based Time Series Model
C. M-MAD: Multi-Modal Anomaly Detection
D. Upgrading Benchmarking with Persistent Anomalies
E. Results and Visualization for M-MAD
IV. CONCLUSION
ACKNOWLEDGMENT
REFERENCES

저자

• Maira Khalid [ Department of AI Convergence Network Ajou University ]
• Ahmed Raza Mohsin [ Department of AI Convergence Network Ajou University ]
• Byeong-hee Roh [ Department of AI Convergence Network Ajou University ] Corresponding Author

참고문헌